BeagleBone AI-64 (4 GB) vs NVIDIA GeForce RTX 4090 for local AI

Pick two to four systems or devices to see which runs more models, faster, for the kind of work you do.

Hardware to compare, two to four

Winner: NVIDIA GeForce RTX 4090

For general chat: Its ten best models average a score of 99 vs 67; it runs 370 more LLMs (496 vs 126); it is about 37.8ร— faster on its top models.

Side by side

Swap sides All models on the BeagleBone AI-64 All models on the RTX 4090

  BeagleBoard BeagleBone AI-64BeagleBone AI-64 (4 GB) NVIDIA GeForce RTX 4090 Founders EditionNVIDIA GeForce RTX 4090
Memory 4.0 GB unified 24.0 GB VRAM
Usable for models 2.8 GB 24.0 GB
Memory bandwidth 13 GB/s 1,008 GB/s
FP16 compute 0.1 TFLOPS 82.6 TFLOPS
Launch price โ€” $1,599
Released 2022 2022
LLMs that run 126 of 622 496 of 622
LLMs that run well 22 466
Image models that run 7 of 61 58 of 61
Video models that run 1 of 43 31 of 43
Decision models that run 41 of 70 68 of 70
Top-10 LLM average score 67 99
Top-10 LLM average speed 2.4 tok/s 89.3 tok/s
Best LLM MiniCPM5 2B Qwen3.8 27B

Highlighted values are the best of the two. Model results assume 8K of context and the default macOS GPU memory limit on Macs.

LLMs

126 of 622 LLMs run on the BeagleBone AI-64 (4 GB). 2.8 GB usable for models.

496 of 622 LLMs run on the NVIDIA GeForce RTX 4090. 24.0 GB usable for models.

Model QualityBeagleBone AI-64 SpeedBeagleBone AI-64 RunsBeagleBone AI-64 QualityRTX 4090 SpeedRTX 4090 RunsRTX 4090
MiniCPM5 2BOpenBMB ยท 2.5B 57Good quality (B) ยท 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 1.7 tok/s Very slowtight fit 60Good quality (B) ยท 60 of 100: how capable the model is at F16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 115 tok/s Runs great
LFM3 1.7BLiquid AI ยท 1.7B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 3.3 tok/s Slow 56Good quality (B) ยท 56 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 168 tok/s Runs great
Spark-X2.6 4BSparkLLM ยท 4.1B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 1.7 tok/s Very slowtight fit 66Good quality (B) ยท 66 of 100: how capable the model is at F16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 70.6 tok/s Runs great
TMax 4BAi2 ยท 4.2B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 1.7 tok/s Very slowtight fit 61Good quality (B) ยท 61 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 69.0 tok/s Runs great
LFM2.5 2.6BLiquid AI ยท 2.7B 41Fair quality (C) ยท 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 1.6 tok/s Very slowtight fit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 108 tok/s Runs great
Jamba Reasoning 3BAI21 Labs ยท 3.2B 41Fair quality (C) ยท 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 1.4 tok/s Very slowtight fit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 90.7 tok/s Runs great
Jamba2 3BAI21 Labs ยท 3.0B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 1.4 tok/s Very slowtight fit 42Fair quality (C) ยท 42 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 95.8 tok/s Runs great
LFM3-VL 2BLiquid AI ยท 1.9B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 50Fair quality (C) ยท 50 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 151 tok/s Runs great
TMax 2BAi2 ยท 1.9B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 3.0 tok/s Slow 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 154 tok/s Runs great
MiniCPM5 1BOpenBMB ยท 1.1B 34Basic quality (D) ยท 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 4.0 tok/s Slow 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 269 tok/s Runs great
LFM2.5-VL 3BLiquid AI ยท 3.1B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowtight fit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 93.0 tok/s Runs great
MiniCPM-V 4.7 1BOpenBMB ยท 1.6B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 3.6 tok/s Slow 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 187 tok/s Runs great
Nemotron 3 Nano 4BNVIDIA ยท 4.0B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowtight fit 50Fair quality (C) ยท 50 of 100: how capable the model is at F16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 73.1 tok/s Runs great
EXAONE 4.0 1.2BLG AI Research ยท 1.3B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 3.4 tok/s Slow 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 227 tok/s Runs great
North Micro VisionCohere ยท 2.5B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 2.9 tok/s Very slowtight fit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 117 tok/s Runs great
Qwen3.5 2BAlibaba ยท 2.3B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slow 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 128 tok/s Runs great
LFM2.5 1.2B InstructLiquid AI ยท 1.2B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 32Basic quality (D) ยท 32 of 100: how capable the model is at F16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 248 tok/s Runs great
LFM2 2.6BLiquid AI ยท 2.6B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 1.4 tok/s Very slowtight fit 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 113 tok/s Runs great
Granite 4.0 H MicroIBM ยท 3.2B 31Basic quality (D) ยท 31 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowtight fit 38Basic quality (D) ยท 38 of 100: how capable the model is at F16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 91.0 tok/s Runs great
MiniCPM-OCR 1.0 1.3BOpenBMB ยท 1.3B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 3.4 tok/s Slow 42Fair quality (C) ยท 42 of 100: how capable the model is at F16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 223 tok/s Runs great
Gemma 3 4B ITGoogle ยท 4.3B 31Basic quality (D) ยท 31 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 1.7 tok/s Very slowtight fit 45Fair quality (C) ยท 45 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 67.5 tok/s Runs great
MedGemma 4B ITGoogle ยท 4.3B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 1.7 tok/s Very slowtight fit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 67.5 tok/s Runs great
Unlimited OCR 2 (3B-A0.6B)Baidu ยท 3.3B (0.6B active) 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 11.5 tok/s Runs welltight fit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 155 tok/s Runs great
PaddleOCR-VL 1.6 0.9BBaidu ยท 1.0B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 4.5 tok/s Slow 42Fair quality (C) ยท 42 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 302 tok/s Runs great
SmolLM3 3BHugging Face ยท 3.1B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 2.3 tok/s Very slowtight fit 42Fair quality (C) ยท 42 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 94.3 tok/s Runs great
Falcon-H1 3B InstructTII ยท 3.2B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowtight fit 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 92.2 tok/s Runs great
TeleOCRXingChen-AGI ยท 1.4B 38Basic quality (D) ยท 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 3.1 tok/s Slowtight fit 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 204 tok/s Runs great
Raon 1 nanoKRAFTON ยท 3.5B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 1.6 tok/s Very slowtight fit 35Basic quality (D) ยท 35 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 82.2 tok/s Runs great
Qwen2.5 3B InstructAlibaba ยท 3.1B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowtight fit 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 93.9 tok/s Runs great
Qwen3-VL 2B InstructAlibaba ยท 2.1B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 3.4 tok/s Slowtight fit 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 136 tok/s Runs great
Unlimited OCR (3B-A0.6B)Baidu ยท 3.3B (0.6B active) 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 11.5 tok/s Runs welltight fit 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 155 tok/s Runs great
LFM2 1.2BLiquid AI ยท 1.2B 24Basic quality (D) ยท 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 248 tok/s Runs great
MiniCPM3 4BOpenBMB ยท 4.1B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowtight fit 38Basic quality (D) ยท 38 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 71.3 tok/s Runs great
Falcon-H1 1.5B InstructTII ยท 1.6B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 3.6 tok/s Slow 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 187 tok/s Runs great
EXAONE 3.5 2.4BLG AI Research ยท 2.4B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 2.3 tok/s Very slowtight fit 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 120 tok/s Runs great
Granite 3.2 2B InstructIBM ยท 2.5B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 2.2 tok/s Very slowtight fit 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 115 tok/s Runs great
Granite 3.3 2BIBM ยท 2.5B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 2.2 tok/s Very slowtight fit 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 115 tok/s Runs great
Llama 3.2 1B InstructMeta ยท 1.2B 20Basic quality (D) ยท 20 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 3.5 tok/s Slow 22Basic quality (D) ยท 22 of 100: how capable the model is at F16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 234 tok/s Runs great
Qwen2.5 1.5B InstructAlibaba ยท 1.5B 20Basic quality (D) ยท 20 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 24Basic quality (D) ยท 24 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 189 tok/s Runs great
Granite 4.0 H 1BIBM ยท 1.5B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 3.9 tok/s Slow 28Basic quality (D) ยท 28 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 199 tok/s Runs great
Qwen3.5 0.8BAlibaba ยท 0.9B 24Basic quality (D) ยท 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 5.0 tok/s Slow 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 334 tok/s Runs great
Granite 3.0 2B InstructIBM ยท 2.6B 24Basic quality (D) ยท 24 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 1.7 tok/s Very slowtight fit 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 110 tok/s Runs great
Granite 3.1 2B InstructIBM ยท 2.5B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 2.2 tok/s Very slowtight fit 28Basic quality (D) ยท 28 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 115 tok/s Runs great
Gemma 2 2BGoogle ยท 2.6B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 2.2 tok/s Very slowtight fit 28Basic quality (D) ยท 28 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 111 tok/s Runs great
Qwen2.5-VL 3B InstructAlibaba ยท 3.8B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 1.9 tok/s Very slowtight fit 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 77.4 tok/s Runs great
Qwen3 1.7BAlibaba ยท 2.0B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 3.5 tok/s Slowtight fit 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 143 tok/s Runs great
LFM2-VL 1.6BLiquid AI ยท 1.6B 20Basic quality (D) ยท 20 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 3.6 tok/s Slow 24Basic quality (D) ยท 24 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 184 tok/s Runs great
Hunyuan 1.8B InstructTencent ยท 1.8B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slow 22Basic quality (D) ยท 22 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 162 tok/s Runs great
Granite 3.1 3B-A800M InstructIBM ยท 3.3B (0.8B active) 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 8.7 tok/s Slowtight fit 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 140 tok/s Runs great
Granite 3.0 3B-A800M InstructIBM ยท 3.4B (0.8B active) 15Basic quality (D) ยท 15 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 6.9 tok/s Slowtight fit 18Basic quality (D) ยท 18 of 100: how capable the model is at F16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 140 tok/s Runs great
Qwen2-VL 2B InstructAlibaba ยท 2.2B 16Basic quality (D) ยท 16 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slow 24Basic quality (D) ยท 24 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 131 tok/s Runs great
Qwen3 0.6BAlibaba ยท 0.8B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 7.5 tok/s Slow 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 387 tok/s Runs great
WeVisOne VL 2BTencent ยท 2.4B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 2.9 tok/s Very slowtight fit 27Basic quality (D) ยท 27 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 119 tok/s Runs great
Hunyuan 0.5B InstructTencent ยท 0.5B 5Basic quality (D) ยท 5 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 13.3 tok/s Runs well 8Basic quality (D) ยท 8 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 538 tok/s Runs great
Granite Vision 3.2 2BIBM ยท 3.0B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 2.4 tok/s Very slowtight fit 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 97.4 tok/s Runs great
Zamba2 2.7B InstructZyphra ยท 2.7B 15Basic quality (D) ยท 15 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 2.7 tok/s Very slow 22Basic quality (D) ยท 22 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 109 tok/s Runs great
Qwen2.5 Coder 3B InstructAlibaba ยท 3.1B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowtight fit 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 93.9 tok/s Runs great
Gemma 3 1BGoogle ยท 1.0B 19Basic quality (D) ยท 19 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 4.4 tok/s Slow 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 290 tok/s Runs great
GLiDE 2BFastino ยท 1.9B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 3.0 tok/s Slow 17Basic quality (D) ยท 17 of 100: how capable the model is at F16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 154 tok/s Runs great
Qwen2 1.5B InstructAlibaba ยท 1.5B 13Basic quality (D) ยท 13 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 16Basic quality (D) ยท 16 of 100: how capable the model is at F16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 189 tok/s Runs great
Falcon 3 1B InstructTII ยท 1.7B 13Basic quality (D) ยท 13 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 3.4 tok/s Slow 16Basic quality (D) ยท 16 of 100: how capable the model is at F16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 174 tok/s Runs great
SmolLM2 135M InstructHugging Face ยท 0.1B 3Basic quality (D) ยท 3 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great 4Basic quality (D) ยท 4 of 100: how capable the model is at F16, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 2,233 tok/s Runs great
Spark X2.6 1.7BSparkLLM ยท 1.7B 13Basic quality (D) ยท 13 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 3.3 tok/s Slow 16Basic quality (D) ยท 16 of 100: how capable the model is at F16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 170 tok/s Runs great
OLMo 2 1B InstructAi2 ยท 1.5B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 3.8 tok/s Slow 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 196 tok/s Runs great
LFM2 700MLiquid AI ยท 0.7B 17Basic quality (D) ยท 17 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 5.9 tok/s Slow 18Basic quality (D) ยท 18 of 100: how capable the model is at F16, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 392 tok/s Runs great
LFM2.5 230MLiquid AI ยท 0.2B 7Basic quality (D) ยท 7 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 31.2 tok/s Runs great 10Basic quality (D) ยท 10 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 1,262 tok/s Runs great
SmolLM 1.7B InstructHugging Face ยท 1.7B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 3.3 tok/s Slow 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 170 tok/s Runs great
Falcon-H1 Tiny 90M InstructTII ยท 0.1B 2Basic quality (D) ยท 2 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great 3Basic quality (D) ยท 3 of 100: how capable the model is at F16, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 3,226 tok/s Runs great
Nemotron Labs Dory 1.0 4BNVIDIA ยท 4.0B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 1.5 tok/s Very slowover memory limit 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 72.9 tok/s Runs great
Qwen2.5 Coder 1.5B InstructAlibaba ยท 1.5B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 189 tok/s Runs great
InternLM2.5 1.8B ChatShanghai AI Lab ยท 1.9B 12Basic quality (D) ยท 12 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 3.8 tok/s Slowtight fit 18Basic quality (D) ยท 18 of 100: how capable the model is at F16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 154 tok/s Runs great
DeepScaleR 1.5B PreviewAgentica ยท 1.8B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slow 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 163 tok/s Runs great
LFM2.5 350MLiquid AI ยท 0.4B 9Basic quality (D) ยท 9 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 20.5 tok/s Runs well 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 829 tok/s Runs great
SmolLM2 360M InstructHugging Face ยท 0.4B 5Basic quality (D) ยท 5 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 19.9 tok/s Runs well 8Basic quality (D) ยท 8 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 806 tok/s Runs great
SmolVLM 256M InstructHugging Face ยท 0.3B 4Basic quality (D) ยท 4 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 27.6 tok/s Runs well 6Basic quality (D) ยท 6 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 1,117 tok/s Runs great
SmolLM 135M InstructHugging Face ยท 0.1B 1Basic quality (D) ยท 1 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great 2Basic quality (D) ยท 2 of 100: how capable the model is at F16, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking. 2,233 tok/s Runs great
Gemma 4 E2BGoogle ยท 5.1B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 2.6 tok/s Very slowover memory limit 62Good quality (B) ยท 62 of 100: how capable the model is at F16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 126 tok/s Runs great
LFM2 350MLiquid AI ยท 0.4B 8Basic quality (D) ยท 8 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 20.5 tok/s Runs well 12Basic quality (D) ยท 12 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 829 tok/s Runs great
SmolVLM 500M InstructHugging Face ยท 0.5B 7Basic quality (D) ยท 7 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 14.1 tok/s Runs well 10Basic quality (D) ยท 10 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 569 tok/s Runs great
StableLM 2 1.6B ChatStability AI ยท 1.6B 7Basic quality (D) ยท 7 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 4.4 tok/s Slow 10Basic quality (D) ยท 10 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 177 tok/s Runs great
StableLM 2 Zephyr 1.6BStability AI ยท 1.6B 7Basic quality (D) ยท 7 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 4.4 tok/s Slow 10Basic quality (D) ยท 10 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 177 tok/s Runs great
Gemma 2B ITGoogle ยท 2.5B 10Basic quality (D) ยท 10 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 2.2 tok/s Very slow 12Basic quality (D) ยท 12 of 100: how capable the model is at F16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. 116 tok/s Runs great
Falcon-H1 0.5B InstructTII ยท 0.5B 9Basic quality (D) ยท 9 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 13.8 tok/s Runs well 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 558 tok/s Runs great
Granite 4.0 H 350MIBM ยท 0.3B 8Basic quality (D) ยท 8 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 21.1 tok/s Runs well 12Basic quality (D) ยท 12 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 854 tok/s Runs great
Phi-2 2.7BMicrosoft ยท 2.8B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 2.6 tok/s Very slowtight fit 16Basic quality (D) ยท 16 of 100: how capable the model is at F16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 104 tok/s Runs great
DeepSeek R1 Distill Qwen 1.5BDeepSeek ยท 1.8B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slow 22Basic quality (D) ยท 22 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 163 tok/s Runs great
LFM2-VL 450MLiquid AI ยท 0.5B 8Basic quality (D) ยท 8 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 15.9 tok/s Runs well 12Basic quality (D) ยท 12 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 645 tok/s Runs great
Qwen3.5 4BAlibaba ยท 4.7B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 1.3 tok/s Very slowover memory limit 62Good quality (B) ยท 62 of 100: how capable the model is at F16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 62.3 tok/s Runs great
Gemma 3 270MGoogle ยท 0.3B 4Basic quality (D) ยท 4 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 26.5 tok/s Runs well 6Basic quality (D) ยท 6 of 100: how capable the model is at F16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 1,075 tok/s Runs great
ERNIE 4.5 0.3BBaidu ยท 0.4B 4Basic quality (D) ยท 4 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 19.9 tok/s Runs well 6Basic quality (D) ยท 6 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 806 tok/s Runs great
Agents-A1 4BInternScience ยท 4.5B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 1.3 tok/s Very slowover memory limit 60Good quality (B) ยท 60 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 63.9 tok/s Runs great
Qwen2.5 0.5B InstructAlibaba ยท 0.5B 7Basic quality (D) ยท 7 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 14.6 tok/s Runs well 10Basic quality (D) ยท 10 of 100: how capable the model is at F16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 592 tok/s Runs great
Qwen2.5-Math 1.5B InstructAlibaba ยท 1.5B 16Basic quality (D) ยท 16 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 189 tok/s Runs great
Fara 1.5 4BMicrosoft ยท 4.5B 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 1.3 tok/s Very slowover memory limit 58Good quality (B) ยท 58 of 100: how capable the model is at F16, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 63.9 tok/s Runs great
GLiDE 0.8BFastino ยท 0.8B 7Basic quality (D) ยท 7 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 9.6 tok/s Slow 11Basic quality (D) ยท 11 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 387 tok/s Runs great
SmolLM 360M InstructHugging Face ยท 0.4B 3Basic quality (D) ยท 3 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 19.9 tok/s Runs well 5Basic quality (D) ยท 5 of 100: how capable the model is at F16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 806 tok/s Runs great
Qwen2 0.5B InstructAlibaba ยท 0.5B 4Basic quality (D) ยท 4 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 14.6 tok/s Runs well 6Basic quality (D) ยท 6 of 100: how capable the model is at F16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 592 tok/s Runs great
d1 2BLiquid AI ยท 1.9B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 3.7 tok/s Slow 21Basic quality (D) ยท 21 of 100: how capable the model is at F16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 151 tok/s Runs great
Granite 4.2 3BIBM ยท 3.7B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 1.7 tok/s Very slowover memory limit 52Fair quality (C) ยท 52 of 100: how capable the model is at F16, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 79.3 tok/s Runs great
StarCoder2 3BBigCode ยท 3.0B 12Basic quality (D) ยท 12 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 2.4 tok/s Very slow 18Basic quality (D) ยท 18 of 100: how capable the model is at F16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 95.8 tok/s Runs great
CodeGemma 2BGoogle ยท 2.5B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 2.2 tok/s Very slow 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 116 tok/s Runs great
TinyLlama 1.1B ChatTinyLlama ยท 1.1B 4Basic quality (D) ยท 4 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 6.5 tok/s Slow 6Basic quality (D) ยท 6 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 264 tok/s Runs great
Granite 4.1 3BIBM ยท 3.4B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowover memory limit 46Fair quality (C) ยท 46 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 85.4 tok/s Runs great
surya ocr 2.5Datalab ยท 0.7B 10Basic quality (D) ยท 10 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking. 10.4 tok/s Runs well 15Basic quality (D) ยท 15 of 100: how capable the model is at F16, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 421 tok/s Runs great
Qwen2.5 Coder 0.5B InstructAlibaba ยท 0.5B 5Basic quality (D) ยท 5 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking. 14.6 tok/s Runs well 8Basic quality (D) ยท 8 of 100: how capable the model is at F16, whatever it runs on. Match score 46/100 (quality, fit, speed and memory use) sets the ranking. 592 tok/s Runs great
Granite 4.0 Micro (3B)IBM ยท 3.4B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking. 1.8 tok/s Very slowover memory limit 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 85.4 tok/s Runs great
S5 nanoMaritaca AI ยท 4.7B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. 1.3 tok/s Very slowover memory limit 37Basic quality (D) ยท 37 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 62.3 tok/s Runs great
Llama 3.2 3B InstructMeta ยท 3.2B 24Basic quality (D) ยท 24 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. 1.9 tok/s Very slowover memory limit 35Basic quality (D) ยท 35 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 90.4 tok/s Runs great
Hermes 3 Llama 3.2 3BNous Research ยท 3.2B 22Basic quality (D) ยท 22 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 43/100 (quality, fit, speed and memory use) sets the ranking. 1.9 tok/s Very slowover memory limit 32Basic quality (D) ยท 32 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 90.4 tok/s Runs great
LightOnOCR-3 4BLightOn ยท 4.5B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking. 1.3 tok/s Very slowover memory limit 48Fair quality (C) ยท 48 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 63.9 tok/s Runs great
Falcon 3 3B InstructTII ยท 3.2B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking. 1.9 tok/s Very slowover memory limit 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 89.9 tok/s Runs great
Phi-4 Mini (3.8B)Microsoft ยท 3.8B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 40/100 (quality, fit, speed and memory use) sets the ranking. 1.6 tok/s Very slowfar over memory limit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 75.6 tok/s Runs great
Ministral 3 3BMistral AI ยท 3.9B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 40/100 (quality, fit, speed and memory use) sets the ranking. 1.6 tok/s Very slowfar over memory limit 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 75.4 tok/s Runs great
Nemotron Mini 4B InstructNVIDIA ยท 4.2B 15Basic quality (D) ยท 15 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 39/100 (quality, fit, speed and memory use) sets the ranking. 1.5 tok/s Very slowover memory limit 22Basic quality (D) ยท 22 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 69.3 tok/s Runs great
Gemma 3n E2BGoogle ยท 5.4B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 39/100 (quality, fit, speed and memory use) sets the ranking. 3.0 tok/s Slowfar over memory limit 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 145 tok/s Runs great
AFM 4.5BArcee AI ยท 4.6B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 39/100 (quality, fit, speed and memory use) sets the ranking. 1.3 tok/s Very slowfar over memory limit 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 62.8 tok/s Runs great
Chandra OCR 2Datalab ยท 5.3B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 36/100 (quality, fit, speed and memory use) sets the ranking. 1.1 tok/s Very slowfar over memory limit 48Fair quality (C) ยท 48 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 54.8 tok/s Runs great
Phi-4 Mini Reasoning (3.8B)Microsoft ยท 3.8B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 36/100 (quality, fit, speed and memory use) sets the ranking. 1.6 tok/s Very slowfar over memory limit 46Fair quality (C) ยท 46 of 100: how capable the model is at F16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 75.6 tok/s Runs great
Qwen2.5-Omni 3BAlibaba ยท 5.5B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 35/100 (quality, fit, speed and memory use) sets the ranking. 1.1 tok/s Very slowfar over memory limit 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 52.4 tok/s Runs great
SmolVLM2 2.2B InstructHugging Face ยท 2.3B 16Basic quality (D) ยท 16 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 35/100 (quality, fit, speed and memory use) sets the ranking. 2.7 tok/s Very slowfar over memory limit 24Basic quality (D) ยท 24 of 100: how capable the model is at F16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 129 tok/s Runs great
Yi Coder 1.5B Chat01.AI ยท 1.5B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 34/100 (quality, fit, speed and memory use) sets the ranking. 4.1 tok/s Slowover memory limit 16Basic quality (D) ยท 16 of 100: how capable the model is at F16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 196 tok/s Runs great
SmolVLM 2.2B InstructHugging Face ยท 2.3B 15Basic quality (D) ยท 15 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 34/100 (quality, fit, speed and memory use) sets the ranking. 2.7 tok/s Very slowfar over memory limit 22Basic quality (D) ยท 22 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 129 tok/s Runs great
SmolLM2 1.7B InstructHugging Face ยท 1.7B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 34/100 (quality, fit, speed and memory use) sets the ranking. 3.6 tok/s Slowfar over memory limit 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 170 tok/s Runs great
DeepSeek Coder 1.3B InstructDeepSeek ยท 1.4B 8Basic quality (D) ยท 8 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 33/100 (quality, fit, speed and memory use) sets the ranking. 4.5 tok/s Slowover memory limit 12Basic quality (D) ยท 12 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 215 tok/s Runs great
LLM4Decompile 1.3B V2LLM4Binary ยท 1.4B 8Basic quality (D) ยท 8 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 33/100 (quality, fit, speed and memory use) sets the ranking. 4.5 tok/s Slowover memory limit 12Basic quality (D) ยท 12 of 100: how capable the model is at F16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 215 tok/s Runs great
StableLM Zephyr 3BStability AI ยท 2.8B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 32/100 (quality, fit, speed and memory use) sets the ranking. 2.2 tok/s Very slowfar over memory limit 16Basic quality (D) ยท 16 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 104 tok/s Runs great
gpt-oss 120BOpenAI ยท 117B (5.1B active) 84Great quality (A) ยท 84 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
gpt-oss 20BOpenAI ยท 21B (3.6B active) 70Great quality (A) ยท 70 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 164 tok/s Runs great
Kimi K3.1 Pro (2.8T)Moonshot AI ยท 2,780B (104.0B active) 69Good quality (B) ยท 69 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 69Good quality (B) ยท 69 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V4.1 Flash 1017 (552B-A16B)DeepSeek ยท 763B (16.0B active) 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-5.3 (753B-A40B)Zhipu AI ยท 753B (40.0B active) 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3.8 Max (2.4T-A95B)Alibaba ยท 2,446B (95.0B active) 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kimi K3 (2.8T)Moonshot AI ยท 2,780B (104.0B active) 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
LongCat 2.5 (1.6T-A48B)Meituan ยท 1,780B (48.0B active) 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiMo V2.6 Pro (1T-A42B)Xiaomi ยท 1,024B (42.0B active) 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Ling 3.1 Flash 1019 (560B-A25B)Ant Group (inclusionAI) ยท 572B (25.0B active) 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax M3.1 Flash (428B-A23B)MiniMax ยท 429B (23.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Nex N2.6 (321B-A18B)Nex-AGI ยท 321B (18.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Step 5 Preview (600B-A27B)StepFun ยท 604B (27.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-5.3 Flash (320B-A18B)Zhipu AI ยท 321B (18.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiMo V2.6 Flash (310B-A15B)Xiaomi ยท 311B (15.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V4.1 Flash (552B-A16B)DeepSeek ยท 763B (16.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Mistral Large 4 (1T-A52B)Mistral AI ยท 1,053B (52.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3.8 Flash Next (125B-A6B)Alibaba ยท 180B (6.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-5.2 (753B-A40B)Zhipu AI ยท 753B (40.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Hunyuan Hy4 Preview (780B)Tencent ยท 780B (55.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Muse Spark Open (816B-A42B)Meta ยท 822B (42.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V4 Pro (1.6T-A49B)DeepSeek ยท 1,651B (49.0B active) 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kimi K2.6 (1T)Moonshot AI ยท 1,027B (32.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Nanbeige4.3 33BBOSS Zhipin (Nanbeige) ยท 33B 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 90Excellent quality (A+) ยท 90 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. 37.4 tok/s Runs welltight fit
Ling 3.1 Flash (560B-A25B)Ant Group (inclusionAI) ยท 572B (25.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiMo V2.5 Pro (1T-A42B)Xiaomi ยท 1,023B (42.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3.8 27BAlibaba ยท 28B 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 92Excellent quality (A+) ยท 92 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. 36.8 tok/s Runs welltight fit
Intern-S2 (397B-A17B)Shanghai AI Lab ยท 403B (17.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-5.1 (744B-A40B)Zhipu AI ยท 754B (40.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Nex N2 Pro (397B-A17B)Nex-AGI ยท 397B (17.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V4 Pro (April 2026)DeepSeek ยท 1,599B (49.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Naive N0.5 Flash (309B-A15.5B)Naive AI ยท 309B (15.5B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3.5 397B-A17BAlibaba ยท 403B (17.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax M3 (427B)MiniMax ยท 427B (25.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-5 (744B-A40B)Zhipu AI ยท 754B (40.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
K2-Horizon (375B-A23B)Institute of Foundation Models ยท 379B (23.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Hunyuan Hy3 (295B-A21B)Tencent ยท 299B (21.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Beam 1.0 (501B-A23B)Reflection AI ยท 501B (23.4B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kimi K2.5 (1T)Moonshot AI ยท 1,027B (32.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kimi K2.7 Code (1T)Moonshot AI ยท 1,027B (32.0B active) 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V4 Flash (284B-A13B)DeepSeek ยท 304B (13.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiMo V2.5 (310B-A15B)Xiaomi ยท 311B (15.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Ornith 1.5 397B-A17BDeepReinforce ยท 403B (17.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3.6 27BAlibaba ยท 28B 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 89Excellent quality (A+) ยท 89 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. 44.5 tok/s Runs great
Nemotron 3.5 Super VL 120B-A12BNVIDIA ยท 124B (12.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V3.2 (671B)DeepSeek ยท 685B (37.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Inkling Small (276B-A12B)Thinking Machines Lab ยท 266B (12.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-4.7 (355B-A32B)Zhipu AI ยท 358B (32.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Inkling (975B-A41B)Thinking Machines Lab ยท 952B (41.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax M2.7 (230B-A10B)MiniMax ยท 229B (10.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3.6 35B-A3BAlibaba ยท 36B (3.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Step 3.7 Flash (198B-A11B)StepFun ยท 201B (11.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniCPM-V 4.7 35B-A3BOpenBMB ยท 35B (3.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Nemotron 3 Ultra 550B-A55BNVIDIA ยท 561B (55.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek R1 0528 (671B)DeepSeek ยท 685B (37.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kimi K2 Thinking (1T)Moonshot AI ยท 1,026B (32.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3.5 27BAlibaba ยท 28B 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. 44.5 tok/s Runs great
Qwen3.5 35B-A3BAlibaba ยท 36B (3.0B active) 65Good quality (B) ยท 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Ornith 1.5 122B-A10BDeepReinforce ยท 125B (10.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V3.2 ExpDeepSeek ยท 685B (37.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Gemma 4 26B-A4BGoogle ยท 26B (3.8B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 91Excellent quality (A+) ยท 91 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 100/100 (quality, fit, speed and memory use) sets the ranking. 124 tok/s Runs welltight fit
Ling 2.6 1TAnt Group (inclusionAI) ยท 1,025B (63.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax M2.5 (230B-A10B)MiniMax ยท 229B (10.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Nemotron 3.5 Bolt 74B-A9BNVIDIA ยท 75B (9.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 41.9 tok/s Runs welloffloads to RAM
DeepSeek R1 (671B)DeepSeek ยท 671B (37.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V4 Flash (April 2026)DeepSeek ยท 291B (13.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Muse Glimmer 30BMeta ยท 30B 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 87Excellent quality (A+) ยท 87 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 99/100 (quality, fit, speed and memory use) sets the ranking. 41.6 tok/s Runs great
Nex N2 Mini (35B-A3B)Nex-AGI ยท 35B (3.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 87Excellent quality (A+) ยท 87 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Qwen3.5 122B-A10BAlibaba ยท 125B (10.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Step 3.5 Flash (196B-A11B)StepFun ยท 199B (11.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V3.1 TerminusDeepSeek ยท 685B (37.0B active) 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V3.1 (671B)DeepSeek ยท 685B (37.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Gemma 4 31BGoogle ยท 31B 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. 39.6 tok/s Runs welltight fit
KAT-Coder V2.5 Dev (35B-A3B)Kuaishou ยท 35B (3.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Kimi K2 Instruct 0905 (1T)Moonshot AI ยท 1,026B (32.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kolibri 1 (78B-A3.5B)Aleph Alpha ยท 78B (3.5B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 83.9 tok/s Runs welloffloads to RAM
MiniMax M2.1 (230B-A10B)MiniMax ยท 229B (10.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3 235B-A22B Thinking 2507Alibaba ยท 235B (22.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3 Coder 480B-A35BAlibaba ยท 480B (35.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Sage-A1.5 35B-A3BInternScience ยท 36B (3.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
MiMo V2 Flash (309B-A15B)Xiaomi ยท 310B (15.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Flint (40B-A2.5B)Recursive ยท 40B (2.5B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 75Great quality (A) ยท 75 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 208 tok/s Runs welltight fit
Ling 1TAnt Group (inclusionAI) ยท 1,000B (50.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Mistral Large 3 (675B)Mistral AI ยท 675B (41.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Ornith 1.5 35B-A3BDeepReinforce ยท 36B (3.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 85Excellent quality (A+) ยท 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Qwen3 235B-A22B Instruct 2507Alibaba ยท 235B (22.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Solar Open 2 250BUpstage ยท 250B (15.0B active) 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
TMax 27BAi2 ยท 27B 63Good quality (B) ยท 63 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 85Excellent quality (A+) ยท 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 98/100 (quality, fit, speed and memory use) sets the ranking. 45.2 tok/s Runs great
Apodex 1.1 Mini (35B-A3B)Apodex ยท 36B (3.0B active) 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Kimi K2 Instruct (July 2025)Moonshot AI ยท 1,026B (32.0B active) 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax M2 (230B-A10B)MiniMax ยท 229B (10.0B active) 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Ornith 1.0 35B-A3BDeepReinforce ยท 36B (3.0B active) 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Qwen3-VL 235B-A22B InstructAlibaba ยท 236B (22.0B active) 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Command A+ (218B-A25B)Cohere ยท 219B (25.0B active) 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Fara 1.5 27BMicrosoft ยท 27B 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. 45.2 tok/s Runs great
GLM-4.7 Flash (30B-A3B)Zhipu AI ยท 31B (3.0B active) 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
K-EXAONE 236B-A23BLG AI Research ยท 237B (23.0B active) 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
LongCat Flash Chat (560B)Meituan ยท 562B (27.0B active) 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Nemotron 3 Super 120B-A12BNVIDIA ยท 124B (12.0B active) 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3 235B-A22BAlibaba ยท 235B (22.0B active) 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Raon-1 mini Preview (30B-A4B)KRAFTON ยท 29B (4.3B active) 61Good quality (B) ยท 61 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. 146 tok/s Runs great
Agents-A1 35B-A3BInternScience ยท 35B (3.0B active) 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 82Great quality (A) ยท 82 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Qwen AgentWorld 35B-A3BAlibaba ยท 35B (3.0B active) 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 82Great quality (A) ยท 82 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Devstral 2 (123B)Mistral AI ยท 125B 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
InternVL3.5 241B-A28BShanghai AI Lab ยท 241B (28.0B active) 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Laguna S 2.1 (118B-A8B)Poolside ยท 118B (8.0B active) 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Llama 3.1 Nemotron Ultra 253B v1NVIDIA ยท 253B 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3 Coder Next (80B-A3B)Alibaba ยท 80B (3.0B active) 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 60.0 tok/s Slowheavy RAM offload
Qwen3-Next 80B-A3B ThinkingAlibaba ยท 81B (3.0B active) 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 58.9 tok/s Slowheavy RAM offload
UI-Mate 27BTencent ยท 27B 60Good quality (B) ยท 60 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 81Great quality (A) ยท 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. 45.2 tok/s Runs great
Hunyuan Hy-Memory 30B-A3BTencent ยท 30B (3.0B active) 59Good quality (B) ยท 59 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 82Great quality (A) ยท 82 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 155 tok/s Runs welltight fit
Baichuan M3 235B-A22BBaichuan ยท 235B (22.0B active) 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-4.6 (355B-A32B)Zhipu AI ยท 357B (32.0B active) 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3 Next 80B-A3B InstructAlibaba ยท 81B (3.0B active) 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 58.9 tok/s Slowheavy RAM offload
DeepSeek V3 0324 (671B)DeepSeek ยท 685B (37.0B active) 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Gemma 4 12BGoogle ยท 12B 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 96/100 (quality, fit, speed and memory use) sets the ranking. 52.7 tok/s Runs great
GLM-4.6VZhipu AI ยท 108B (12.0B active) 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax-M1 80KMiniMax ยท 456B (45.9B active) 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Command A Reasoning (111B)Cohere ยท 111B 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
EXAONE 4.5 33BLG AI Research ยท 34B 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 77Great quality (A) ยท 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 36.0 tok/s Runs welltight fit
GLM-4.5 (355B-A32B)Zhipu AI ยท 358B (32.0B active) 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Laguna XS 2.1 (33B-A3B)Poolside ยท 33B (3.0B active) 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 77Great quality (A) ยท 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Ling 3.0 Flash (127B-A5B)Ant Group (inclusionAI) ยท 127B (5.1B active) 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax-M1 40KMiniMax ยท 456B (45.9B active) 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Mistral Small 4 (119B-A6B)Mistral AI ยท 119B (6.0B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Mistral Medium 3.5 (128B)Mistral AI ยท 128B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Trinity Large Thinking (400B-A13B)Arcee AI ยท 399B (13.0B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Command A (111B)Cohere ยท 111B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
ERNIE 4.5 300B-A47BBaidu ยท 300B (47.0B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
GLM-4.5VZhipu AI ยท 108B (12.0B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Hermes 4 405BNous Research ยท 406B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
K2 Horizon 7BInstitute of Foundation Models ยท 9.0B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 70.0 tok/s Runs great
Kimi Dev 72BMoonshot AI ยท 73B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 5.0 tok/s Slowheavy RAM offload
Ling 2.6 Flash (104B-A7.4B)Ant Group (inclusionAI) ยท 107B (7.4B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Llama 3.3 Nemotron Super 49B v1.5NVIDIA ยท 50B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 31.2 tok/s Runs welloffloads to RAM
Mistral Large 2.1 (2411)Mistral AI ยท 123B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Nemotron Cascade 2 30B-A3BNVIDIA ยท 32B (3.5B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 76Great quality (A) ยท 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 161 tok/s Runs welltight fit
North Mini Code 1.0 (30B-A3B)Cohere ยท 30B (3.0B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 77Great quality (A) ยท 77 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 155 tok/s Runs welltight fit
Seed-OSS 36B InstructByteDance ยท 36B 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 44.2 tok/s Runs welltight fit
Solar Open 100BUpstage ยท 103B (12.0B active) 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 19.8 tok/s Slowheavy RAM offload
Hermes 4.3 36BNous Research ยท 36B 55Good quality (B) ยท 55 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 44.2 tok/s Runs welltight fit
DiffusionGemma 26B-A4B ITGoogle ยท 26B (4.0B active) 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 77Great quality (A) ยท 77 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 121 tok/s Runs welltight fit
GLM-4.5 Air (106B-A12B)Zhipu AI ยท 110B (12.0B active) 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiMo V2.6 Distill Qwen 9BXiaomi ยท 9.4B 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 78Great quality (A) ยท 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 66.9 tok/s Runs great
Mistral Large 2 (123B)Mistral AI ยท 123B 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
NorthCUA Probeipo-ai ยท 397B (17.0B active) 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Pixtral Large 124B (2411)Mistral AI ยท 124B 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3 30B-A3B Thinking 2507Alibaba ยท 31B (3.3B active) 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 166 tok/s Runs welltight fit
Qwen3 Coder 30B-A3BAlibaba ยท 31B (3.3B active) 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 166 tok/s Runs welltight fit
Qwen3-VL 32B InstructAlibaba ยท 33B 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 48.0 tok/s Runs great
QwQ 32BAlibaba ยท 33B 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 37.8 tok/s Runs welltight fit
Ring Flash 2.0 (100B-A6B)Ant Group (inclusionAI) ยท 103B (6.1B active) 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 38.2 tok/s Slowheavy RAM offload
SenseNova Loop 9B PreviewSenseTime ยท 9.7B 54Fair quality (C) ยท 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 78Great quality (A) ยท 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 65.2 tok/s Runs great
Nemotron 3 Nano 30B-A3BNVIDIA ยท 32B (3.5B active) 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 73Great quality (A) ยท 73 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 161 tok/s Runs welltight fit
Ornith 1.5 9BDeepReinforce ยท 9.7B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 77Great quality (A) ยท 77 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 65.3 tok/s Runs great
Qwen3 30B-A3B Instruct 2507Alibaba ยท 31B (3.3B active) 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 73Great quality (A) ยท 73 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 166 tok/s Runs welltight fit
Devstral Small 2 (24B)Mistral AI ยท 24B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 73Great quality (A) ยท 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
dots.llm1 Instructrednote ยท 143B (14.0B active) 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Granite 4.2 30BIBM ยท 29B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 42.3 tok/s Runs welltight fit
InternVL3.5 38BShanghai AI Lab ยท 38B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 41.7 tok/s Runs welltight fit
Ling Flash 2.0 (100B-A6B)Ant Group (inclusionAI) ยท 103B (6.1B active) 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 38.2 tok/s Slowheavy RAM offload
MiniMax-Text-01MiniMax ยท 456B (45.9B active) 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Olmo 3.1 32B ThinkAi2 ยท 32B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 38.4 tok/s Runs welltight fit
OpenReasoning Nemotron 32BNVIDIA ยท 33B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 37.8 tok/s Runs welltight fit
Qwen3 32BAlibaba ยท 33B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 37.7 tok/s Runs welltight fit
Qwen3-VL 30B-A3B ThinkingAlibaba ยท 31B (3.3B active) 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 166 tok/s Runs welltight fit
Qwen3.5 9BAlibaba ยท 9.7B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 76Great quality (A) ยท 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 65.3 tok/s Runs great
Skywork OR1 32BSkywork ยท 33B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 37.8 tok/s Runs welltight fit
Tรผlu 3 405BAi2 ยท 406B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V3 (671B)DeepSeek ยท 671B (37.0B active) 52Fair quality (C) ยท 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Ornith 1.0 9BDeepReinforce ยท 9.7B 52Fair quality (C) ยท 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 75Great quality (A) ยท 75 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 65.3 tok/s Runs great
Sarvam 105BSarvam AI ยท 106B (10.0B active) 52Fair quality (C) ยท 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. 24.0 tok/s Slowheavy RAM offload
DeepSeek Coder V2 (236B)DeepSeek ยท 236B (21.0B active) 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
EXAONE Deep 32BLG AI Research ยท 32B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 38.7 tok/s Runs welltight fit
GLM-4 32B 0414Zhipu AI ยท 33B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 38.0 tok/s Runs welltight fit
GLM-Z1 32B (0414)Zhipu AI ยท 33B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 38.0 tok/s Runs welltight fit
Hermes 4 70BNous Research ยท 71B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
Hunyuan A13B (80B-A13B)Tencent ยท 80B (13.0B active) 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 15.5 tok/s Slowheavy RAM offload
InternVL3 78BShanghai AI Lab ยท 78B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 4.1 tok/s Slowheavy RAM offload
Kimi Linear 48B-A3BMoonshot AI ยท 49B (3.0B active) 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 217 tok/s Runs welltight fit
Leanstral 119B-A6BMistral AI ยท 119B (6.5B active) 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
LFM2.5 24B-A2BLiquid AI ยท 24B (2.5B active) 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 71Great quality (A) ยท 71 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 170 tok/s Runs great
Llama 3.3 70B InstructMeta ยท 71B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
Qwen3 30B-A3BAlibaba ยท 31B (3.3B active) 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 166 tok/s Runs welltight fit
Qwen3-VL 30B-A3B InstructAlibaba ยท 31B (3.3B active) 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 166 tok/s Runs welltight fit
TMax 9BAi2 ยท 9.0B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 70.4 tok/s Runs great
Qwen2.5 72B InstructAlibaba ยท 73B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 5.0 tok/s Slowheavy RAM offload
Qwen2.5-VL 72B InstructAlibaba ยท 73B 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 51Fair quality (C) ยท 51 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 4.8 tok/s Slowheavy RAM offload
Apriel 1.5 15B ThinkerServiceNow ยท 15B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 42.4 tok/s Runs great
DeepSeek Coder V2 Instruct (0614)DeepSeek ยท 236B (21.0B active) 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek R1 Distill Llama 70BDeepSeek ยท 71B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
DeepSeek R1 Distill Qwen 32BDeepSeek ยท 33B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 37.7 tok/s Runs welltight fit
DeepSeek V2.5 1210DeepSeek ยท 236B (21.0B active) 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
EXAONE 4.0 32BLG AI Research ยท 32B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 38.7 tok/s Runs welltight fit
Gemma 3 27B ITGoogle ยท 27B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 45.2 tok/s Runs great
Granite 4.1 30BIBM ยท 29B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 42.9 tok/s Runs welltight fit
Llama 3.1 405B InstructMeta ยท 406B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Llama 4 Maverick (17B x 128E)Meta ยท 402B (17.0B active) 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Llama 4 Scout 17B-16EMeta ยท 109B (17.0B active) 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Magistral Small 1.2 (24B)Mistral AI ยท 24B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 69Good quality (B) ยท 69 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Nemotron 3.5 Lightning 30B-A3BNVIDIA ยท 32B (3.0B active) 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Olmo 3.1 32B InstructAi2 ยท 32B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 38.4 tok/s Runs welltight fit
Phi-4 Reasoning Plus (14B)Microsoft ยท 15B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 43.0 tok/s Runs great
Tรผlu 3 70BAi2 ยท 71B 50Fair quality (C) ยท 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
Baichuan M2 32BBaichuan ยท 33B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 37.8 tok/s Runs welltight fit
Command R+ (08-2024) 104BCohere ยท 104B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepSeek V2.5DeepSeek ยท 236B (21.0B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Devstral Small 1.1 (24B)Mistral AI ยท 24B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great
ERNIE 4.5 21B-A3B ThinkingBaidu ยท 22B (3.0B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 69Good quality (B) ยท 69 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 141 tok/s Runs great
Groot 65B A6B 261023GrootTeam ยท 118B (6.0B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
InternVL3 38BShanghai AI Lab ยท 38B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 41.7 tok/s Runs welltight fit
Jamba 1.5 LargeAI21 Labs ยท 399B (94.0B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Llama 3.1 70B InstructMeta ยท 71B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
Magistral Small 2507Mistral AI ยท 24B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great
Maple Preview (20B-A1B)DeepGrove ยท 20B (1.5B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 69Good quality (B) ยท 69 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 187 tok/s Runs great
MedGemma 27B Text ITGoogle ยท 27B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 45.8 tok/s Runs great
Mellum2.1 Thinking (12B-A2.5B)JetBrains ยท 12B (2.5B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 129 tok/s Runs great
Nemotron 3.5 Bolt E4BNVIDIA ยท 8.0B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 64.5 tok/s Runs great
OpenHands LM 32BAll Hands AI ยท 33B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 37.8 tok/s Runs welltight fit
Phi-4 Reasoning (14B)Microsoft ยท 15B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 43.0 tok/s Runs great
Phi-4 Reasoning Vision 15BMicrosoft ยท 15B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 41.7 tok/s Runs great
Qwen2.5 32B InstructAlibaba ยท 33B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 37.7 tok/s Runs welltight fit
Qwen3-Omni 30B-A3B InstructAlibaba ยท 35B (3.3B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 189 tok/s Runs great
Sarvam 30BSarvam AI ยท 32B (2.4B active) 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 189 tok/s Runs welltight fit
UI-Mate 9BTencent ยท 9.4B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 66.9 tok/s Runs great
UI-Venus 2 9BAnt Group (inclusionAI) ยท 9.4B 49Fair quality (C) ยท 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 66.9 tok/s Runs great
Gemma 4 E4BGoogle ยท 8.0B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 64.5 tok/s Runs great
Granite 4.2 8BIBM ยท 8.8B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 71.7 tok/s Runs great
Hermes 3 (Llama 3.1 70B)Nous Research ยท 71B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
Ling 3.0 flash SanteAnt Group (inclusionAI) ยท 127B (5.1B active) 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Magistral Small 2506Mistral AI ยท 24B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great
Mistral Small 3.2 24BMistral AI ยท 24B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Nanbeige4.2 3BBOSS Zhipin (Nanbeige) ยท 4.2B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 69.6 tok/s Runs great
QVQ 72B PreviewAlibaba ยท 73B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 4.8 tok/s Slowheavy RAM offload
Qwen2.5-VL 32B InstructAlibaba ยท 33B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 47.8 tok/s Runs great
Qwen3 14BAlibaba ยท 15B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Terminal RSI 27BTeriminal-RSI ยท 28B 47Fair quality (C) ยท 47 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 44.5 tok/s Runs great
Hemmingway 1.5Hemmingway ยท 27B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 46.0 tok/s Runs great
Qwen2.5 Coder 32B InstructAlibaba ยท 33B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 37.7 tok/s Runs welltight fit
AceReason Nemotron 14BNVIDIA ยท 15B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Apertus v1.5 70BSwiss AI ยท 72B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 55Good quality (B) ยท 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 3.0 tok/s Slowheavy RAM offload
Apriel Nemotron 15B ThinkerServiceNow ยท 15B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 42.1 tok/s Runs great
Devstral Small 2505Mistral AI ยท 24B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great
Hermes 4 14BNous Research ยท 15B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Llama 3.1 Nemotron 70B InstructNVIDIA ยท 71B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 55Good quality (B) ยท 55 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
Llama 3.2 90B Vision InstructMeta ยท 89B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 3.0 tok/s Slowheavy RAM offload
Ministral 3 14B ReasoningMistral AI ยท 14B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 45.2 tok/s Runs great
Mistral Small 3.1 24BMistral AI ยท 24B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
OpenReasoning Nemotron 14BNVIDIA ยท 15B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Qwen2.5 Coder 14B InstructAlibaba ยท 15B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
QwQ 32B PreviewAlibaba ยท 33B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 37.8 tok/s Runs welltight fit
Reka Flash 3.1 (21B)Reka AI ยท 21B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 41.1 tok/s Runs great
SenseNova 6.8 Code2VideoSenseTime ยท 27B 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 45.2 tok/s Runs great
Trinity Mini (26B-A3B)Arcee AI ยท 26B (3.0B active) 46Fair quality (C) ยท 46 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 141 tok/s Runs welltight fit
Command R+ 104B (04-2024)Cohere ยท 104B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
DeepCoder 14B PreviewAgentica ยท 15B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
ERNIE 4.5 21B-A3BBaidu ยท 22B (3.0B active) 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 141 tok/s Runs great
Falcon-H1 34B InstructTII ยท 34B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 36.8 tok/s Runs welltight fit
InternVL3 14BShanghai AI Lab ยท 15B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 41.7 tok/s Runs great
Ministral 3 14BMistral AI ยท 14B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 45.2 tok/s Runs great
Mistral Small 3 24B (2501)Mistral AI ยท 24B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great
Phi-4 (14B)Microsoft ยท 15B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 42.9 tok/s Runs great
Qwen2 72B InstructAlibaba ยท 73B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 5.0 tok/s Slowheavy RAM offload
Qwen2-VL 72B InstructAlibaba ยท 73B 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 4.8 tok/s Slowheavy RAM offload
Axion 1.0 35BAxion ยท 35B (3.0B active) 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
Codestral 22BMistral AI ยท 22B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 45.9 tok/s Runs great
Command R (08-2024) 32BCohere ยท 32B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 38.3 tok/s Runs welltight fit
DeepSeek R1 0528 Qwen3 8BDeepSeek ยท 8.2B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 76.9 tok/s Runs great
DeepSeek R1 Distill Qwen 14BDeepSeek ยท 15B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
DeepSeek V2 ChatDeepSeek ยท 236B (21.0B active) 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Dolphin Mistral 24B Venice EditionCognitive Computations ยท 24B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Gemma 2 27B ITGoogle ยท 27B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 45.5 tok/s Runs great
Granite 4.0 H Small (32B-A9B)IBM ยท 32B (9.0B active) 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 94.0 tok/s Runs welltight fit
InternVL3.5 8BShanghai AI Lab ยท 8.5B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 73.8 tok/s Runs great
JanusCoder 14BShanghai AI Lab ยท 15B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Qwen2.5 14B InstructAlibaba ยท 15B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 42.9 tok/s Runs great
Qwen3-VL 8B InstructAlibaba ยท 8.8B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 71.8 tok/s Runs great
Voxtral Small 24BMistral AI ยท 24B 43Fair quality (C) ยท 43 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 42.1 tok/s Runs great
Apertus 70B InstructSwiss AI ยท 71B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
Aya Expanse 32BCohere ยท 32B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 38.3 tok/s Runs welltight fit
Aya Vision 32BCohere ยท 33B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 37.3 tok/s Runs welltight fit
Dolphin 3.0 R1 Mistral 24BCognitive Computations ยท 24B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 43.4 tok/s Runs great
Gemma 3 12B ITGoogle ยท 12B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 51.6 tok/s Runs great
GLM-4.1V 9B ThinkingZhipu AI ยท 10B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 61.2 tok/s Runs great
Granite 4.1 8BIBM ยท 8.8B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 71.7 tok/s Runs great
Kimi-VL A3B Thinking 2506Moonshot AI ยท 16B (3.0B active) 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 117 tok/s Runs great
Ling 3.0 Tiny (8B-A1.3B)Ant Group (inclusionAI) ยท 7.9B (1.3B active) 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 113 tok/s Runs great
Mage-VL-Cowork 4.7BMicrosoft ยท 4.7B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at F16, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 61.2 tok/s Runs great
MiniCPM-V 4.5 8BOpenBMB ยท 8.7B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 72.4 tok/s Runs great
Nemotron Nano 9B v2NVIDIA ยท 8.9B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 70.9 tok/s Runs great
OLMo 2 32B InstructAi2 ยท 32B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 57Good quality (B) ยท 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 38.4 tok/s Runs welltight fit
Qwen2.5 14B Instruct 1MAlibaba ยท 15B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 42.6 tok/s Runs great
Qwen3 8BAlibaba ยท 8.2B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 76.9 tok/s Runs great
Solar Pro Preview 22BUpstage ยท 22B 42Fair quality (C) ยท 42 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 46.2 tok/s Runs great
NorthCUA Minibeipo-ai ยท 35B (3.0B active) 41Fair quality (C) ยท 41 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 173 tok/s Runs welltight fit
GLM-Z1 9B (0414)Zhipu AI ยท 9.4B 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 67.0 tok/s Runs great
Llama 3 70B InstructMeta ยท 71B 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 48Fair quality (C) ยท 48 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 3.2 tok/s Slowheavy RAM offload
LLM4Decompile 22B V2LLM4Binary ยท 22B 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 45.9 tok/s Runs great
MiMo-VL 7B RLXiaomi ยท 8.3B 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 75.8 tok/s Runs great
Qwen2.5-Math 72B InstructAlibaba ยท 73B 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 5.6 tok/s Slowoffloads to RAM
Seed-Coder 8B ReasoningByteDance ยท 8.3B 40Fair quality (C) ยท 40 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 76.4 tok/s Runs great
Groot 19B A3B 261023GrootTeam ยท 19B (3.0B active) 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 141 tok/s Runs great
EXAONE Deep 7.8BLG AI Research ยท 7.8B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 80.6 tok/s Runs great
Falcon-H1R 7BTII ยท 7.6B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 83.0 tok/s Runs great
InternVL3 8BShanghai AI Lab ยท 7.9B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 79.3 tok/s Runs great
LFM2 24B-A2BLiquid AI ยท 24B (2.3B active) 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 174 tok/s Runs great
Mage-VL 4.7BMicrosoft ยท 4.7B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 61.2 tok/s Runs great
MiMo 7B RLXiaomi ยท 7.8B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 80.4 tok/s Runs great
Ministral 3 8BMistral AI ยท 8.9B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 70.6 tok/s Runs great
Phi-3.5 MoE (42B-A6.6B)Microsoft ยท 42B (6.6B active) 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 47Fair quality (C) ยท 47 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 136 tok/s Runs welltight fit
Qwen3 4B Thinking 2507Alibaba ยท 4.0B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 72.2 tok/s Runs great
Seed-Coder 8B InstructByteDance ยท 8.3B 39Basic quality (D) ยท 39 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 76.4 tok/s Runs great
Mixtral 8x22B InstructMistral AI ยท 141B (39.0B active) 38Basic quality (D) ยท 38 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen2.5 Coder 7B InstructAlibaba ยท 7.6B 38Basic quality (D) ยท 38 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 55Good quality (B) ยท 55 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 82.7 tok/s Runs great
Command R 35B (v01)Cohere ยท 35B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 16.4 tok/s Runs welloffloads to RAM
Llama 3.1 Nemotron Nano 8B v1NVIDIA ยท 8.0B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
MiniCPM4 8BOpenBMB ยท 8.2B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 76.9 tok/s Runs great
Mistral Small 22B (2409)Mistral AI ยท 22B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 45.9 tok/s Runs great
Olmo 3 7B ThinkAi2 ยท 7.3B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 86.3 tok/s Runs great
OpenReasoning Nemotron 7BNVIDIA ยท 7.6B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 82.7 tok/s Runs great
Qwen3 4B Instruct 2507Alibaba ยท 4.0B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at F16, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 72.2 tok/s Runs great
Qwen3-VL 4B InstructAlibaba ยท 4.4B 37Basic quality (D) ยท 37 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at F16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 65.4 tok/s Runs great
DeepSeek Coder V2 Lite (16B)DeepSeek ยท 16B (2.4B active) 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 131 tok/s Runs great
Falcon 3 10B InstructTII ยท 10B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 61.1 tok/s Runs great
Foundation-Sec 8B ReasoningCisco Foundation AI ยท 8.0B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
GLM-4 9B 0414Zhipu AI ยท 9.4B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 67.0 tok/s Runs great
Hunyuan 7B InstructTencent ยท 7.5B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 84.0 tok/s Runs great
LFM2.5 8B-A1BLiquid AI ยท 8.5B (1.5B active) 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at F16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 105 tok/s Runs great
Ling Mini 2.0 (16B-A1.4B)Ant Group (inclusionAI) ยท 16B (1.4B active) 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 166 tok/s Runs great
MiniCPM-o 2.6 8BOpenBMB ยท 8.7B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 72.7 tok/s Runs great
Mistral Nemo 12B InstructMistral AI ยท 12B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 51.6 tok/s Runs great
NorthCUA Nanobeipo-ai ยท 9.4B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 66.9 tok/s Runs great
Nous Hermes 2 Mixtral 8x7B DPONous Research ยท 47B (12.9B active) 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 106 tok/s Runs welltight fit
Qwen1.5 110B ChatAlibaba ยท 111B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen2.5 7B InstructAlibaba ยท 7.6B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 82.7 tok/s Runs great
Qwen2.5-VL 7B InstructAlibaba ยท 8.3B 36Basic quality (D) ยท 36 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 76.0 tok/s Runs great
Aya 23 35BCohere ยท 35B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 16.4 tok/s Runs welloffloads to RAM
DeepSeek R1 Distill Llama 8BDeepSeek ยท 8.0B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
Falcon-H1 7B InstructTII ยท 7.6B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 83.0 tok/s Runs great
Fara 7BMicrosoft ยท 8.3B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 76.0 tok/s Runs great
Foundation-Sec 1.1 8B InstructCisco Foundation AI ยท 8.0B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
Gemma 2 9B ITGoogle ยท 9.2B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 68.2 tok/s Runs great
GLiDE 9BFastino ยท 7.9B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 79.3 tok/s Runs great
Grok-1 (314B)xAI ยท 314B (86.0B active) 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
InternLM2.5 20B ChatShanghai AI Lab ยท 20B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 49Fair quality (C) ยท 49 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 43.2 tok/s Runs great
InternLM3 8B InstructShanghai AI Lab ยท 8.8B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 71.6 tok/s Runs great
InternVL2.5 8BShanghai AI Lab ยท 8.1B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 78.0 tok/s Runs great
Jamba 1.5 MiniAI21 Labs ยท 52B (12.0B active) 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 111 tok/s Runs welltight fit
Kimi-VL A3B InstructMoonshot AI ยท 16B (3.0B active) 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 117 tok/s Runs great
Mixtral 8x7B InstructMistral AI ยท 47B (12.9B active) 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 106 tok/s Runs welltight fit
Olmo 3 7B InstructAi2 ยท 7.3B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 86.3 tok/s Runs great
Pixtral 12B (2409)Mistral AI ยท 12B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 50.8 tok/s Runs great
Qianfan-OCR 4BBaidu ยท 4.7B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 61.2 tok/s Runs great
Qwen2 57B-A14B InstructAlibaba ยท 57B (14.0B active) 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 65.6 tok/s Runs welloffloads to RAM
Qwen2.5 7B Instruct 1MAlibaba ยท 7.6B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 82.7 tok/s Runs great
Qwen3 4BAlibaba ยท 4.0B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at F16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 72.2 tok/s Runs great
Snowflake Arctic InstructSnowflake ยท 479B (17.0B active) 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Yi 1.5 34B Chat01.AI ยท 34B 35Basic quality (D) ยท 35 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 47Fair quality (C) ยท 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 36.0 tok/s Runs welltight fit
GLM-4 9B Chat (2024)Zhipu AI ยท 9.4B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 48Fair quality (C) ยท 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 67.0 tok/s Runs great
GLM-4 9B Chat 1MZhipu AI ยท 9.5B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 48Fair quality (C) ยท 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 66.4 tok/s Runs great
Hermes 3 (Llama 3.1 8B)Nous Research ยท 8.0B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 48Fair quality (C) ยท 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
Llama 3.1 8B InstructMeta ยท 8.0B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 48Fair quality (C) ยท 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
MiniCPM-V 2.6 8BOpenBMB ยท 8.1B 33Basic quality (D) ยท 33 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 48Fair quality (C) ยท 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 77.8 tok/s Runs great
Apertus v1.5 8BSwiss AI ยท 8.9B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 70.8 tok/s Runs great
CodeGeeX4 All 9BZhipu AI ยท 9.4B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 67.0 tok/s Runs great
DeepHermes 3 Llama 3 8B PreviewNous Research ยท 8.0B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
DeepSeek R1 Distill Qwen 7BDeepSeek ยท 7.6B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 82.7 tok/s Runs great
Dolphin 3.0 Llama 3.1 8BCognitive Computations ยท 8.0B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
EXAONE 3.5 7.8BLG AI Research ยท 7.8B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 80.6 tok/s Runs great
Falcon 3 7B InstructTII ยท 7.5B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 84.4 tok/s Runs great
Gemma 3n E4BGoogle ยท 7.9B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at F16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 72.6 tok/s Runs great
Granite 3.2 8B InstructIBM ยท 8.2B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 77.1 tok/s Runs great
Granite 3.3 8BIBM ยท 8.2B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 77.1 tok/s Runs great
Ministral 8B (2410)Mistral AI ยท 8.0B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 78.5 tok/s Runs great
Moonlight 16B-A3B InstructMoonshot AI ยท 16B (3.0B active) 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 117 tok/s Runs great
Phi-3 Medium (14B)Microsoft ยท 14B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 45.1 tok/s Runs great
Phi-3 Medium 128K InstructMicrosoft ยท 14B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 45.1 tok/s Runs great
Qwen2.5-Omni 7BAlibaba ยท 11B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 58.7 tok/s Runs great
Tรผlu 3 8BAi2 ยท 8.0B 32Basic quality (D) ยท 32 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 46Fair quality (C) ยท 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
WeVisOne-VL 4BTencent ยท 4.8B 31Basic quality (D) ยท 31 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 45Fair quality (C) ยท 45 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 60.1 tok/s Runs great
Aya Vision 8BCohere ยท 8.6B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 73.0 tok/s Runs great
Code Llama 70B InstructMeta ยท 69B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. 6.6 tok/s Slowoffloads to RAM
Command R7BCohere ยท 8.0B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
DeepSeek Coder 33B InstructDeepSeek ยท 33B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 48.0 tok/s Runs great
Dolphin 2.9.4 Llama 3.1 8BCognitive Computations ยท 8.0B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
Granite 3.1 8B InstructIBM ยท 8.2B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 77.1 tok/s Runs great
Granite 4.0 H Tiny (7B-A1B)IBM ยท 6.9B (1.0B active) 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 128 tok/s Runs great
Hermes 2 Pro Llama 3 8BNous Research ยท 8.0B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
InternVL2 8BShanghai AI Lab ยท 8.1B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 78.0 tok/s Runs great
Llama 3.1 Nemotron Nano 4B v1.1NVIDIA ยท 4.5B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 64.4 tok/s Runs great
LLaVA 1.6 34BLLaVA ยท 35B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 46.0 tok/s Runs great
OLMo 2 13B InstructAi2 ยท 14B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 45.9 tok/s Runs great
OpenChat 3.6 8BOpenChat ยท 8.0B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
Phi-3 Small 8K InstructMicrosoft ยท 7.4B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 85.2 tok/s Runs great
Phi-4 Multimodal InstructMicrosoft ยท 5.6B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at F16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 52.1 tok/s Runs great
Qwen1.5 32B ChatAlibaba ยท 33B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 41Fair quality (C) ยท 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 38.1 tok/s Runs welltight fit
Qwen2 7B InstructAlibaba ยท 7.6B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 82.7 tok/s Runs great
Qwen2-VL 7B InstructAlibaba ยท 8.3B 30Basic quality (D) ยท 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 4/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 44Fair quality (C) ยท 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 76.0 tok/s Runs great
Command R7B ArabicCohere ยท 8.0B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 42Fair quality (C) ยท 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
Hunyuan 4B InstructTencent ยท 4.2B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 42Fair quality (C) ยท 42 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 68.8 tok/s Runs great
InternLM2 Chat 20BShanghai AI Lab ยท 20B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 41Fair quality (C) ยท 41 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 43.2 tok/s Runs great
Llama 3 8B InstructMeta ยท 8.0B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 42Fair quality (C) ยท 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
Llama 3.2 11B Vision InstructMeta ยท 11B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 42Fair quality (C) ยท 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 59.0 tok/s Runs great
Mistral-NeMo-Minitron 8B InstructNVIDIA ยท 8.4B 29Basic quality (D) ยท 29 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 42Fair quality (C) ยท 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 74.9 tok/s Runs great
Apertus 8B InstructSwiss AI ยท 8.1B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 78.2 tok/s Runs great
Aya Expanse 8BCohere ยท 8.0B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
DeepSeek LLM 67B ChatDeepSeek ยท 67B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 7.0 tok/s Slowoffloads to RAM
DeepSeek Prover V2 7BDeepSeek ยท 6.9B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 42.0 tok/s Runs great
DeepSeek V2 Lite ChatDeepSeek ยท 16B (2.4B active) 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 131 tok/s Runs great
Granite 3.0 8B InstructIBM ยท 8.2B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 77.1 tok/s Runs great
Granite Code 34B InstructIBM ยท 34B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 36.7 tok/s Runs welltight fit
InternLM2 Math Plus 20BShanghai AI Lab ยท 20B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 39Basic quality (D) ยท 39 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 43.2 tok/s Runs great
InternLM2.5 7B ChatShanghai AI Lab ยท 7.7B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 81.4 tok/s Runs great
Nous Hermes 2 SOLAR 10.7BNous Research ยท 11B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 58.7 tok/s Runs great
Phi-3.5 Mini InstructMicrosoft ยท 3.8B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 76.0 tok/s Runs great
SOLAR 10.7B InstructUpstage ยท 11B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 58.7 tok/s Runs great
StarCoder2 15B InstructBigCode ยท 16B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 39.5 tok/s Runs great
Yi 34B Chat01.AI ยท 34B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 36.0 tok/s Runs welltight fit
Yi Coder 9B Chat01.AI ยท 8.8B 28Basic quality (D) ยท 28 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 40Fair quality (C) ยท 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 71.3 tok/s Runs great
Code Llama 34B InstructMeta ยท 34B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 36.7 tok/s Runs welltight fit
CogVLM2 Llama 3 Chat 19BZhipu AI ยท 20B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 52.4 tok/s Runs great
Dolphin 2.9 Llama 3 8BCognitive Computations ยท 8.0B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
EXAONE 3.0 7.8B InstructLG AI Research ยท 7.8B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 80.6 tok/s Runs great
Falcon3 Mamba 7B InstructTII ยท 7.3B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 86.6 tok/s Runs great
LFM2 8B-A1BLiquid AI ยท 8.3B (1.5B active) 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at F16, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 105 tok/s Runs great
LLM4Decompile 9B V2LLM4Binary ยท 8.8B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 71.3 tok/s Runs great
Mistral 7B Instruct v0.3Mistral AI ยท 7.3B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at F16, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 40.0 tok/s Runs great
Phi-3.5 Vision InstructMicrosoft ยท 4.2B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 70.0 tok/s Runs great
Zamba2 7B InstructZyphra ยท 7.4B 26Basic quality (D) ยท 26 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 38Basic quality (D) ยท 38 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 85.6 tok/s Runs great
Codestral Mamba 7BMistral AI ยท 7.3B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 86.4 tok/s Runs great
Granite Code 20B InstructIBM ยท 20B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 42.8 tok/s Runs great
Hermes 2 Pro Mistral 7BNous Research ยท 7.2B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Llama 2 70B ChatMeta ยท 69B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 47/100 (quality, fit, speed and memory use) sets the ranking. 6.6 tok/s Slowoffloads to RAM
OLMo 2 7B InstructAi2 ยท 7.3B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 86.3 tok/s Runs great
Phi-3 Mini 128K InstructMicrosoft ยท 3.8B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 76.0 tok/s Runs great
Phi-3 Mini 4K InstructMicrosoft ยท 3.8B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 76.0 tok/s Runs great
Qwen2.5-Math 7B InstructAlibaba ยท 7.6B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 82.7 tok/s Runs great
Starling LM 7B BetaNexusflow ยท 7.2B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Voxtral Mini 3BMistral AI ยท 4.7B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 62.0 tok/s Runs great
Yi 1.5 9B Chat01.AI ยท 8.8B 25Basic quality (D) ยท 25 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 36Basic quality (D) ยท 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 71.3 tok/s Runs great
OpenChat 3.5 (0106)OpenChat ยท 7.2B 24Basic quality (D) ยท 24 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 35Basic quality (D) ยท 35 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Aya 23 8BCohere ยท 8.0B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 78.4 tok/s Runs great
CodeGemma 7BGoogle ยท 8.5B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 73.8 tok/s Runs great
DeepSeek Coder 6.7B InstructDeepSeek ยท 6.7B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 43.1 tok/s Runs great
Falcon 180B ChatTII ยท 180B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Mathstral 7BMistral AI ยท 7.3B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 40.0 tok/s Runs great
Mistral 7B Instruct v0.2Mistral AI ยท 7.2B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Nous Hermes 2 Mistral 7B DPONous Research ยท 7.2B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
OpenChat 3.5 (1210)OpenChat ยท 7.2B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Phi-3 Vision 128K InstructMicrosoft ยท 4.2B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 70.0 tok/s Runs great
Qwen1.5 14B ChatAlibaba ยท 14B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 33Basic quality (D) ยท 33 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 60.6 tok/s Runs great
RecurrentGemma 9B ITGoogle ยท 9.6B 23Basic quality (D) ยท 23 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 34Basic quality (D) ยท 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 65.4 tok/s Runs great
Granite Code 8B Instruct (128K)IBM ยท 8.1B 22Basic quality (D) ยท 22 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 32Basic quality (D) ยท 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 78.2 tok/s Runs great
InternLM2 Math Plus 7BShanghai AI Lab ยท 7.7B 22Basic quality (D) ยท 22 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 32Basic quality (D) ยท 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 81.4 tok/s Runs great
LLM4Decompile 6.7B V2LLM4Binary ยท 6.7B 22Basic quality (D) ยท 22 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 32Basic quality (D) ยท 32 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 43.1 tok/s Runs great
OpenHermes 2.5 Mistral 7BTeknium ยท 7.2B 22Basic quality (D) ยท 22 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 32Basic quality (D) ยท 32 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
StarCoder2 15B (base)BigCode ยท 16B 22Basic quality (D) ยท 22 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 32Basic quality (D) ยท 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 39.5 tok/s Runs great
Starling LM 7B AlphaBerkeley ยท 7.2B 22Basic quality (D) ยท 22 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 3/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 32Basic quality (D) ยท 32 of 100: how capable the model is at F16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Code Llama 13B InstructMeta ยท 13B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 29Basic quality (D) ยท 29 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 65.9 tok/s Runs great
Falcon Mamba 7B InstructTII ยท 7.3B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 30Basic quality (D) ยท 30 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 86.6 tok/s Runs great
Mistral 7B Instruct v0.1Mistral AI ยท 7.2B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
SQLCoder 7B v2Defog ยท 6.7B 21Basic quality (D) ยท 21 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 30Basic quality (D) ยท 30 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 43.1 tok/s Runs great
Gemma 1.1 7B ITGoogle ยท 8.5B 19Basic quality (D) ยท 19 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 28Basic quality (D) ยท 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 73.8 tok/s Runs great
LLaVA 1.6 Mistral 7BLLaVA ยท 7.6B 19Basic quality (D) ยท 19 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 28Basic quality (D) ยท 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 83.2 tok/s Runs great
Neural Chat 7B v3.3Intel ยท 7.2B 19Basic quality (D) ยท 19 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 28Basic quality (D) ยท 28 of 100: how capable the model is at F16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Orca 2 13BMicrosoft ยท 13B 19Basic quality (D) ยท 19 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 28Basic quality (D) ยท 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 48.4 tok/s Runs great
StableLM 2 12B ChatStability AI ยท 12B 19Basic quality (D) ยท 19 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 28Basic quality (D) ยท 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 51.9 tok/s Runs great
Yi 1.5 6B Chat01.AI ยท 6.1B 19Basic quality (D) ยท 19 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 28Basic quality (D) ยท 28 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 47.9 tok/s Runs great
Baichuan 2 13B ChatBaichuan ยท 14B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 45.3 tok/s Runs great
DeepSeek MoE 16B ChatDeepSeek ยท 16B (2.8B active) 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 121 tok/s Runs great
Gemma 7B ITGoogle ยท 8.5B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 73.8 tok/s Runs great
Llama 2 13B ChatMeta ยท 13B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 48.4 tok/s Runs great
OLMoE 1B-7B InstructAi2 ยท 6.9B (1.3B active) 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 113 tok/s Runs great
Qwen1.5 7B ChatAlibaba ยท 7.7B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 81.6 tok/s Runs great
StarCoder 15BBigCode ยท 16B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 39.8 tok/s Runs great
WizardMath 7B v1.1WizardLM ยท 7.2B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Zephyr 7B BetaHugging Face ยท 7.2B 18Basic quality (D) ยท 18 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 26Basic quality (D) ยท 26 of 100: how capable the model is at F16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 40.1 tok/s Runs great
Code Llama 7B InstructMeta ยท 6.7B 16Basic quality (D) ยท 16 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 24Basic quality (D) ยท 24 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 43.1 tok/s Runs great
Qwen1.5 MoE A2.7B ChatAlibaba ยท 14B (2.7B active) 16Basic quality (D) ยท 16 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 24Basic quality (D) ยท 24 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 124 tok/s Runs great
StarCoder2 7BBigCode ยท 7.2B 16Basic quality (D) ยท 16 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 24Basic quality (D) ยท 24 of 100: how capable the model is at F16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 40.5 tok/s Runs great
LLaVA 1.5 13BLLaVA ยท 13B 15Basic quality (D) ยท 15 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 22Basic quality (D) ยท 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 47.2 tok/s Runs great
Orca 2 7BMicrosoft ยท 6.7B 15Basic quality (D) ยท 15 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 22Basic quality (D) ยท 22 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 43.1 tok/s Runs great
WizardLM 13B v1.2WizardLM ยท 13B 15Basic quality (D) ยท 15 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 2/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 22Basic quality (D) ยท 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 48.4 tok/s Runs great
Baichuan 2 7B ChatBaichuan ยท 7.5B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 20Basic quality (D) ยท 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 83.9 tok/s Runs great
Granite Code 3B Instruct (128K)IBM ยท 3.5B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 83.4 tok/s Runs great
Llama 2 7B ChatMeta ยท 6.7B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 43.1 tok/s Runs great
Stable Code Instruct 3BStability AI ยท 2.8B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 104 tok/s Runs great
Vicuna 13B v1.5LMSYS ยท 13B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 20Basic quality (D) ยท 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 48.4 tok/s Runs great
Yi 6B Chat01.AI ยท 6.1B 14Basic quality (D) ยท 14 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 20Basic quality (D) ยท 20 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 47.9 tok/s Runs great
DeepSeek LLM 7B ChatDeepSeek ยท 6.9B 12Basic quality (D) ยท 12 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 18Basic quality (D) ยท 18 of 100: how capable the model is at F16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 42.0 tok/s Runs great
Falcon 40B InstructTII ยท 42B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 11Basic quality (D) ยท 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 48.6 tok/s Runs great
LLaVA 1.5 7BLLaVA ยท 7.1B 11Basic quality (D) ยท 11 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 16Basic quality (D) ยท 16 of 100: how capable the model is at F16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 41.1 tok/s Runs great
OLMo 7B InstructAi2 ยท 6.9B 9Basic quality (D) ยท 9 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 42.1 tok/s Runs great
Vicuna 7B v1.5LMSYS ยท 6.7B 9Basic quality (D) ยท 9 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 14Basic quality (D) ยท 14 of 100: how capable the model is at F16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 43.1 tok/s Runs great
InternLM 7BShanghai AI Lab ยท 7.3B 8Basic quality (D) ยท 8 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 1/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 12Basic quality (D) ยท 12 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 86.1 tok/s Runs great
Falcon 7B InstructTII ยท 7.2B 4Basic quality (D) ยท 4 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 0/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 6Basic quality (D) ยท 6 of 100: how capable the model is at F16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 40.2 tok/s Runs great

Image models

7 of 61 image models run on the BeagleBone AI-64 (4 GB). 2.8 GB usable for models.

58 of 61 image models run on the NVIDIA GeForce RTX 4090. 24.0 GB usable for models.

Model QualityBeagleBone AI-64 SpeedBeagleBone AI-64 RunsBeagleBone AI-64 QualityRTX 4090 SpeedRTX 4090 RunsRTX 4090
SD TurboStability AI ยท 0.9B 28Basic quality (D) ยท 28 of 100: how capable the model is at FP16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 30 s Runs welltight fit ยท encoder unloaded after use 28Basic quality (D) ยท 28 of 100: how capable the model is at FP16, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. 0.1 s Runs great
Realistic Vision V5.1SG161222 ยท 0.9B 42Fair quality (C) ยท 42 of 100: how capable the model is at FP16, whatever it runs on. Match score 54/100 (quality, fit, speed and memory use) sets the ranking. 7.6 min Very slowtight fit ยท encoder unloaded after use 42Fair quality (C) ยท 42 of 100: how capable the model is at FP16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 1.2 s Runs great
DreamShaper 8Lykon ยท 0.9B 40Fair quality (C) ยท 40 of 100: how capable the model is at FP16, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. 7.6 min Very slowtight fit ยท encoder unloaded after use 40Fair quality (C) ยท 40 of 100: how capable the model is at FP16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 1.2 s Runs great
LCM DreamShaper v7Tsinghua (LCM) ยท 0.9B 30Basic quality (D) ยท 30 of 100: how capable the model is at FP16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 1.6 min Slowtight fit ยท encoder unloaded after use 30Basic quality (D) ยท 30 of 100: how capable the model is at FP16, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 0.3 s Runs great
PixArt-ฮฃ XL 1024PixArt ยท 0.6B 50Fair quality (C) ยท 50 of 100: how capable the model is at FP16, whatever it runs on. Match score 51/100 (quality, fit, speed and memory use) sets the ranking. 11.8 min Very slowCPU offload 50Fair quality (C) ยท 50 of 100: how capable the model is at FP16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 1.5 s Runs great
Stable Diffusion 1.5Stability AI / Runway ยท 0.9B 35Basic quality (D) ยท 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 50/100 (quality, fit, speed and memory use) sets the ranking. 7.6 min Very slowtight fit ยท encoder unloaded after use 35Basic quality (D) ยท 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. 1.2 s Runs great
SDXL TurboStability AI ยท 2.6B 39Basic quality (D) ยท 39 of 100: how capable the model is at FP8, whatever it runs on. Match score 48/100 (quality, fit, speed and memory use) sets the ranking. 63 s SlowCPU offload 40Fair quality (C) ยท 40 of 100: how capable the model is at FP16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 0.2 s Runs great
Fibo 2.0 DevBria AI ยท 8.1B 93Excellent quality (A+) ยท 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 93Excellent quality (A+) ยท 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 24 s Runs wellencoder in RAM
MiniMax H3 ImageMiniMax ยท 33.1B 91Excellent quality (A+) ยท 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 91Excellent quality (A+) ยท 91 of 100: how capable the model is at FP8, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 12 s Runs greatCPU offload
Krea 2 TurboKrea ยท 12.8B 90Excellent quality (A+) ยท 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 90Excellent quality (A+) ยท 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking. 3.9 s Runs great
Cosmos3-Super Text2Image (64B)NVIDIA ยท 64.0B 90Excellent quality (A+) ยท 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 90Excellent quality (A+) ยท 90 of 100: how capable the model is at FP8, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
HunyuanImage 3.0 (80B-A13B)Tencent ยท 83.0B 88Excellent quality (A+) ยท 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen-Image 2.1Alibaba ยท 7.1B 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen-ImageAlibaba ยท 20.4B 86Excellent quality (A+) ยท 86 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 86Excellent quality (A+) ยท 86 of 100: how capable the model is at FP8, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 1.8 min SlowCPU offload
FLUX.2 [dev]Black Forest Labs ยท 32.2B 85Excellent quality (A+) ยท 85 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 85Excellent quality (A+) ยท 85 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 45 s Slowtight fit ยท encoder in RAM
Ideogram 4Ideogram ยท 18.6B 85Excellent quality (A+) ยท 85 of 100: how capable the model is at NF4, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 85Excellent quality (A+) ยท 85 of 100: how capable the model is at NF4, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 28 s Runs welltight fit
FLUX 3 Image KleinBlack Forest Labs ยท 5.6B 84Great quality (A) ยท 84 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 90Excellent quality (A+) ยท 90 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 6.6 s Runs great
Ming-Image 0.1 DesignAnt Group (inclusionAI) ยท 6.2B 84Great quality (A) ยท 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 15 s Runs welltight fit ยท encoder in RAM
Qwen-Image 2512Alibaba ยท 20.4B 83Great quality (A) ยท 83 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 1.9 min Slowencoder in RAM
Qwen-Image 2.1 TurboAlibaba ยท 7.1B 82Great quality (A) ยท 82 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 87Excellent quality (A+) ยท 87 of 100: how capable the model is at FP8, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 11 s Runs welltight fit ยท encoder in RAM
ERNIE-ImageBaidu ยท 8.0B 81Great quality (A) ยท 81 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 86Excellent quality (A+) ยท 86 of 100: how capable the model is at FP8, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 20 s Runs well
FireRed-Image-Edit 1.1rednote (FireRed Team) ยท 20.4B 80Great quality (A) ยท 80 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 1.7 min Slowencoder in RAM
Qwen-Image-Edit 2511Alibaba ยท 20.4B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 79Great quality (A) ยท 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 1.7 min Slowencoder in RAM
ERNIE-Image TurboBaidu ยท 8.0B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 3.7 s Runs great
FLUX.2 [klein] 9BBlack Forest Labs ยท 9.1B 78Great quality (A) ยท 78 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 82Great quality (A) ยท 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 1.6 s Runs great
HiDream-I1 FullHiDream.ai ยท 17.1B 78Great quality (A) ยท 78 of 100: how capable the model is at NF4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 82Great quality (A) ยท 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 43 s Slowencoder in RAM
Z-Image TurboAlibaba (Tongyi) ยท 6.2B 78Great quality (A) ยท 78 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 2.9 s Runs great
FLUX.1 Kontext [dev]Black Forest Labs ยท 11.9B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at FP8, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 20 s Runs well
FLUX.1 Krea [dev]Black Forest Labs ยท 11.9B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at FP8, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 11 s Runs great
MOSS-Image SFTOpenMOSS ยท 5.1B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at FP8, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 13 s Runs great
Z-ImageAlibaba (Tongyi) ยท 6.2B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 15 s Runs well
FLUX.1 Fill [dev]Black Forest Labs ยท 11.9B 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 11 s Runs great
FLUX.1 [dev]Black Forest Labs ยท 11.9B 74Great quality (A) ยท 74 of 100: how capable the model is at NF4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 78Great quality (A) ยท 78 of 100: how capable the model is at FP8, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 11 s Runs great
HiDream-I1 DevHiDream.ai ยท 17.1B 74Great quality (A) ยท 74 of 100: how capable the model is at NF4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 78Great quality (A) ยท 78 of 100: how capable the model is at FP8, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 16 s Runs wellencoder in RAM
Chroma1-HDLodestone Rock ยท 8.9B 72Great quality (A) ยท 72 of 100: how capable the model is at NF4, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 76Great quality (A) ยท 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 18 s Runs well
FLUX.2 [klein] 4BBlack Forest Labs ยท 3.9B 72Great quality (A) ยท 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 78Great quality (A) ยท 78 of 100: how capable the model is at FP16, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 1.0 s Runs great
Iris-3BSperidlabs ยท 3.0B 72Great quality (A) ยท 72 of 100: how capable the model is at FP16, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at FP16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 16 s Runs well
HiDream-I1 FastHiDream.ai ยท 17.1B 70Great quality (A) ยท 70 of 100: how capable the model is at NF4, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 9.5 s Runs greatencoder in RAM
CogView4 6BZhipu AI ยท 6.4B 68Good quality (B) ยท 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 16 s Runs well
FLUX.1 [schnell]Black Forest Labs ยท 11.9B 65Good quality (B) ยท 65 of 100: how capable the model is at NF4, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 1.8 s Runs great
OmniGen2BAAI (VectorSpaceLab) ยท 4.0B 64Good quality (B) ยท 64 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at FP8, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 14 s Runs well
Stable Diffusion 3.5 LargeStability AI ยท 8.1B 61Good quality (B) ยท 61 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at FP8, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 9.9 s Runs great
Juggernaut XL v9RunDiffusion ยท 2.6B 60Good quality (B) ยท 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at FP16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 4.5 s Runs great
Lumina-Image 2.0Shanghai AI Lab ยท 2.6B 60Good quality (B) ยท 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at FP16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 7.0 s Runs great
RealVisXL V5.0SG161222 ยท 2.6B 60Good quality (B) ยท 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at FP16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 4.5 s Runs great
Animagine XL 4.0Cagliostro Lab ยท 2.6B 58Good quality (B) ยท 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 4.2 s Runs great
Illustrious XL v0.1OnomaAI ยท 2.6B 58Good quality (B) ยท 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 4.2 s Runs great
KolorsKuaishou ยท 2.6B 58Good quality (B) ยท 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 4.0 s Runs great
Playground v2.5Playground AI ยท 2.6B 58Good quality (B) ยท 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 7.5 s Runs great
Stable Diffusion 3.5 Large TurboStability AI ยท 8.1B 57Good quality (B) ยท 57 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 61Good quality (B) ยท 61 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 1.8 s Runs great
DreamShaper XL v2 TurboLykon ยท 2.6B 56Good quality (B) ยท 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 1.3 s Runs great
Pony Diffusion V6 XLPurpleSmartAI ยท 2.6B 56Good quality (B) ยท 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 3.8 s Runs great
Stable Diffusion 3.5 MediumStability AI ยท 2.5B 56Good quality (B) ยท 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 6.0 s Runs great
Animagine XL 3.1Cagliostro Lab ยท 2.6B 54Fair quality (C) ยท 54 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at FP16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 4.2 s Runs great
AuraFlow v0.3fal ยท 6.9B 53Fair quality (C) ยท 53 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 13 s Runs great
Stable Diffusion XL 1.0Stability AI ยท 2.6B 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 55Good quality (B) ยท 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 4.5 s Runs great
SDXL-Lightning (4-step)ByteDance ยท 2.6B 52Fair quality (C) ยท 52 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 54Fair quality (C) ยท 54 of 100: how capable the model is at FP16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 0.7 s Runs great
Hunyuan-DiT v1.2Tencent ยท 1.5B 52Fair quality (C) ยท 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 6.0 s Runs great
Stable CascadeStability AI ยท 5.2B 50Fair quality (C) ยท 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 5.0 s Runs great
Stable Diffusion 3 MediumStability AI ยท 2.0B 50Fair quality (C) ยท 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 3.1 s Runs great
Kandinsky 3Sber AI ยท 3.0B 48Fair quality (C) ยท 48 of 100: how capable the model is at FP16, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 48Fair quality (C) ยท 48 of 100: how capable the model is at FP16, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 9.5 s Runs greatencoder in RAM

Video models

1 of 43 video models run on the BeagleBone AI-64 (4 GB). 2.8 GB usable for models.

31 of 43 video models run on the NVIDIA GeForce RTX 4090. 24.0 GB usable for models.

Model QualityBeagleBone AI-64 SpeedBeagleBone AI-64 RunsBeagleBone AI-64 QualityRTX 4090 SpeedRTX 4090 RunsRTX 4090
AnimateDiff v3 (SD 1.5)Shanghai AI Lab ยท 1.3B 35Basic quality (D) ยท 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking. 2.6 h Very slowtight fit ยท CPU offload 35Basic quality (D) ยท 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. 20 s Runs great
Cosmos3-Super Image2Video (64B)NVIDIA ยท 64.0B 88Excellent quality (A+) ยท 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kandinsky 6.0 Pro (29B, video + audio)Sber AI ยท 30.1B 88Excellent quality (A+) ยท 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
MiniMax H3 (33B, video + audio)MiniMax ยท 20.1B 83Great quality (A) ยท 83 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at FP8, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 20.3 min Slowtight fit ยท CPU offload
Bernini-R A14B (video generation + editing)ByteDance ยท 14.3B 83Great quality (A) ยท 83 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at FP8, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 11.3 min Runs wellencoder in RAM
HunyuanVideo 1.5 (8.3B)Tencent ยท 8.3B 82Great quality (A) ยท 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 82Great quality (A) ยท 82 of 100: how capable the model is at FP8, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
LTX-2.5 (22B, video + audio)Lightricks ยท 21.0B 81Great quality (A) ยท 81 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 87Excellent quality (A+) ยท 87 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 2.3 min Runs greatCPU offload
LTX-2.3Lightricks ยท 18.9B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 7.8 min Runs wellCPU offload
Wan 2.2 I2V A14BAlibaba ยท 14.0B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 31.2 min Slowtight fit ยท CPU offload
Wan 2.2 T2V A14BAlibaba ยท 14.0B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at FP8, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 31.2 min Slowtight fit ยท CPU offload
SCAIL-2 14B (character animation)Zhipu AI ยท 16.4B 78Great quality (A) ยท 78 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at FP16, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 26.0 min SlowCPU offload
SkyReels V2 T2V 14B 720PSkywork ยท 14.0B 76Great quality (A) ยท 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 76Great quality (A) ยท 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
LTX-2 (19B, video + audio)Lightricks ยท 18.9B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at FP8, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 7.8 min Runs wellCPU offload
HY-OmniWeaving (8.3B, unified video)Tencent ยท 8.3B 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 8.3 min Runs wellencoder in RAM
Step-Video T2V (30B)StepFun ยท 30.0B 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Wan-Dancer 14B (music to dance)Alibaba ยท 17.3B 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at FP8, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Wan 2.1 I2V 14B 720PAlibaba ยท 14.0B 72Great quality (A) ยท 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 76Great quality (A) ยท 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 39.0 min Slowtight fit ยท CPU offload
Wan 2.1 T2V 14BAlibaba ยท 14.0B 72Great quality (A) ยท 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 76Great quality (A) ยท 76 of 100: how capable the model is at FP8, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. 39.0 min Slowtight fit ยท CPU offload
Wan 2.1 VACE 14BAlibaba ยท 17.3B 72Great quality (A) ยท 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 57.2 min Very slowtight fit ยท CPU offload
Wan 2.2 Animate 14BAlibaba ยท 17.3B 72Great quality (A) ยท 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 49.4 min Very slowtight fit ยท CPU offload
FramePack I2V (HunyuanVideo 13B)lllyasviel ยท 12.9B 72Great quality (A) ยท 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 72Great quality (A) ยท 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 6.3 min Runs wellencoder in RAM
SkyReels V2 T2V 14B 540PSkywork ยท 14.0B 72Great quality (A) ยท 72 of 100: how capable the model is at FP8, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 74Great quality (A) ยท 74 of 100: how capable the model is at FP16, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 30.3 min SlowCPU offload
Wan 2.2 S2V 14B (speech to video)Alibaba ยท 16.3B 70Great quality (A) ยท 70 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 20.0 min Slowtight fit
Open-Sora 2.0 (11B)HPC-AI Tech ยท 11.0B 70Great quality (A) ยท 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. 65.0 min Very slowCPU offload
Wan 2.1 I2V 14B 480PAlibaba ยท 14.0B 68Good quality (B) ยท 68 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 18.0 min Slowtight fit
Cosmos-Predict2 14B Video2WorldNVIDIA ยท 14.0B 68Good quality (B) ยท 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Kandinsky 6.0 Video Lite (3B, video + audio)Sber AI ยท 3.2B 68Good quality (B) ยท 68 of 100: how capable the model is at FP16, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 5.4 min Runs wellencoder in RAM
HunyuanVideo 13BTencent ยท 13.0B 66Good quality (B) ยท 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
HunyuanVideo I2V 13BTencent ยท 13.0B 66Good quality (B) ยท 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
LTX-Video 13B 0.9.8 (distilled)Lightricks ยท 13.0B 66Good quality (B) ยท 66 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 66Good quality (B) ยท 66 of 100: how capable the model is at FP8, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 70 s Runs greatCPU offload
Wan 2.2 TI2V 5BAlibaba ยท 5.0B 65Good quality (B) ยท 65 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 68Good quality (B) ยท 68 of 100: how capable the model is at FP8, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 10.1 min Runs welltight fit ยท CPU offload
Mochi 1 Preview (10B)Genmo ยท 10.0B 60Good quality (B) ยท 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 17.3 min Slowtight fit ยท encoder in RAM
Cosmos 1.0 Diffusion 7B Text2WorldNVIDIA ยท 7.0B 60Good quality (B) ยท 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 60Good quality (B) ยท 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
CogVideoX 1.5 5BZhipu AI ยท 5.6B 58Good quality (B) ยท 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 58Good quality (B) ยท 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Cosmos-Predict2 2B Video2WorldNVIDIA ยท 2.0B 56Good quality (B) ยท 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 56Good quality (B) ยท 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 3.9 min Runs welltight fit ยท encoder in RAM
CogVideoX 5BZhipu AI ยท 5.0B 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 2.7 min Runs great
CogVideoX 5B I2VZhipu AI ยท 5.6B 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 2.7 min Runs great
Wan 2.1 T2V 1.3BAlibaba ยท 1.3B 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 3.6 min Runs great
Wan 2.1 VACE 1.3BAlibaba ยท 2.2B 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 53Fair quality (C) ยท 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 3.9 min Runs great
Allegro (2.8B)Rhymes AI ยท 2.8B 52Fair quality (C) ยท 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 52Fair quality (C) ยท 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 7/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
SkyReels V2 Diffusion Forcing 1.3BSkywork ยท 1.4B 50Fair quality (C) ยท 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 4.5 min Runs well
LTX-Video 2BLightricks ยท 1.9B 49Fair quality (C) ยท 49 of 100: how capable the model is at FP8, whatever it runs on. Match score 6/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 50Fair quality (C) ยท 50 of 100: how capable the model is at FP16, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 25 s Runs great
CogVideoX 2BZhipu AI ยท 1.7B 44Fair quality (C) ยท 44 of 100: how capable the model is at FP8, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 45Fair quality (C) ยท 45 of 100: how capable the model is at FP16, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 60 s Runs great

Decision models

41 of 70 decision models run on the BeagleBone AI-64 (4 GB). 2.8 GB usable for models.

68 of 70 decision models run on the NVIDIA GeForce RTX 4090. 24.0 GB usable for models.

Model QualityBeagleBone AI-64 SpeedBeagleBone AI-64 RunsBeagleBone AI-64 QualityRTX 4090 SpeedRTX 4090 RunsRTX 4090
Jina Reranker v3.5Jina AI ยท 600M 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. 25.9 s Very slow 86Excellent quality (A+) ยท 86 of 100: how capable the model is at F16, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 11 ms Runs great
Jina Reranker v3Jina AI ยท 600M 84Great quality (A) ยท 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 25.9 s Very slow 84Great quality (A) ยท 84 of 100: how capable the model is at F16, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 11 ms Runs great
GLiNER2.5 DecideFastino ยท 490M 80Great quality (A) ยท 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 11.4 s Very slow 80Great quality (A) ยท 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 6.6 ms Runs great
mxbai-rerank-large v2Mixedbread ยท 1.5B 81Great quality (A) ยท 81 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 66.2 s Very slow 82Great quality (A) ยท 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 23 ms Runs great
FRIDA-DecisionsSber AI ยท 820M 80Great quality (A) ยท 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 19.0 s Very slow 80Great quality (A) ยท 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 8.4 ms Runs great
Qwen3 Reranker 0.6BAlibaba ยท 600M 80Great quality (A) ยท 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 25.9 s Very slow 80Great quality (A) ยท 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 11 ms Runs great
OpenAI Privacy FilterOpenAI ยท 1.4B (50M active) 78Great quality (A) ยท 78 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. 1.2 s Slowtight fit 78Great quality (A) ยท 78 of 100: how capable the model is at F16, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 4.3 ms Runs great
Llama Prompt Guard 2 86MMeta ยท 280M 76Great quality (A) ยท 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 13.0 s Very slow 76Great quality (A) ยท 76 of 100: how capable the model is at F16, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 7.0 ms Runs great
zerank-1 smallZeroEntropy ยท 1.7B 77Great quality (A) ยท 77 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 73.9 s Very slow 80Great quality (A) ยท 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 25 ms Runs great
mxbai-rerank-base v2Mixedbread ยท 490M 76Great quality (A) ยท 76 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. 21.1 s Very slow 76Great quality (A) ยท 76 of 100: how capable the model is at F16, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 9.9 ms Runs great
GLiClass ModernBERT Large v3Knowledgator ยท 400M 74Great quality (A) ยท 74 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 9.3 s Very slow 74Great quality (A) ยท 74 of 100: how capable the model is at F16, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 6.2 ms Runs great
d1 3BLiquid AI ยท 3.1B 77Great quality (A) ยท 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 67.3 s Very slowtight fit 82Great quality (A) ยท 82 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 28 ms Runs great
DeBERTa v3 Large Zero-shot v2.0Moritz Laurer ยท 440M 72Great quality (A) ยท 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 10.2 s Very slow 72Great quality (A) ยท 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 6.4 ms Runs great
BGE Reranker v2 GemmaBAAI ยท 2.5B 75Great quality (A) ยท 75 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 107.8 s Very slowtight fit 78Great quality (A) ยท 78 of 100: how capable the model is at F16, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 34 ms Runs great
DuoGuard 0.5BDuoGuard ยท 490M 72Great quality (A) ยท 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 22.7 s Very slow 72Great quality (A) ยท 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 9.3 ms Runs great
BGE Reranker v2 M3BAAI ยท 570M 72Great quality (A) ยท 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 26.4 s Very slow 72Great quality (A) ยท 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 10 ms Runs great
GTE Reranker ModernBERT BaseAlibaba ยท 150M 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 7.0 s Very slow 70Great quality (A) ยท 70 of 100: how capable the model is at F16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 5.6 ms Runs great
Llama Prompt Guard 2 22MMeta ยท 71M 68Good quality (B) ยท 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 3.3 s Very slow 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 4.8 ms Runs great
ModernBERT Large Zero-shot v2.0Moritz Laurer ยท 400M 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 9.3 s Very slow 70Great quality (A) ยท 70 of 100: how capable the model is at F16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 6.2 ms Runs great
Skywork Reward V2 Qwen3 0.6BSkywork ยท 600M 72Great quality (A) ยท 72 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 55.7 s Very slow 72Great quality (A) ยท 72 of 100: how capable the model is at F16, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 17 ms Runs great
Jina Reranker v2 Base MultilingualJina AI ยท 280M 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 13.0 s Very slow 70Great quality (A) ยท 70 of 100: how capable the model is at F16, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 7.0 ms Runs great
Granite Embedding Reranker English R2IBM ยท 150M 68Good quality (B) ยท 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 7.0 s Very slow 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 5.6 ms Runs great
GLiGuard 300MFastino ยท 210M 68Good quality (B) ยท 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 9.7 s Very slow 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 6.3 ms Runs great
LayaConvai Innovations ยท 420M 68Good quality (B) ยท 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 9.7 s Very slow 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 6.3 ms Runs great
GLiClass Base v3Knowledgator ยท 190M 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 4.4 s Very slow 66Good quality (B) ยท 66 of 100: how capable the model is at F16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 5.0 ms Runs great
ProtectAI DeBERTa v3 Prompt Injection v2Protect AI ยท 180M 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 8.4 s Very slow 66Good quality (B) ยท 66 of 100: how capable the model is at F16, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 5.9 ms Runs great
DeBERTa v3 Base Zero-shot v2.0Moritz Laurer ยท 180M 64Good quality (B) ยท 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 4.2 s Very slow 64Good quality (B) ยท 64 of 100: how capable the model is at F16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 5.0 ms Runs great
Qwen3Guard Gen 0.6BAlibaba ยท 750M 74Great quality (A) ยท 74 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 34.0 s Very slow 74Great quality (A) ยท 74 of 100: how capable the model is at F16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 26 ms Runs great
GLiNER2 Base v1Fastino ยท 210M 68Good quality (B) ยท 68 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. 4.9 s Very slow 68Good quality (B) ยท 68 of 100: how capable the model is at F16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 5.1 ms Runs great
HHEM 2.1 OpenVectara ยท 110M 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. 10.2 s Very slow 62Good quality (B) ยท 62 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 6.4 ms Runs great
NuExtract 2.0 2BNuMind ยท 2.2B 69Good quality (B) ยท 69 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 64.7 s Very slowtight fit 70Great quality (A) ยท 70 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. 98 ms Runs great
Laya MultilingualConvai Innovations ยท 320M 60Good quality (B) ยท 60 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 7.4 s Very slow 60Good quality (B) ยท 60 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 5.7 ms Runs great
LettuceDetect LargeKR Labs ยท 400M 66Good quality (B) ยท 66 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 37.1 s Very slow 66Good quality (B) ยท 66 of 100: how capable the model is at F16, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. 13 ms Runs great
ms-marco MiniLM L6 v2Sentence Transformers ยท 23M 52Fair quality (C) ยท 52 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 1.1 s Slow 52Fair quality (C) ยท 52 of 100: how capable the model is at F16, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. 4.2 ms Runs great
GLiNER Multi v2.1Urchade Zaratiana ยท 290M 62Good quality (B) ยท 62 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. 6.7 s Very slow 62Good quality (B) ยท 62 of 100: how capable the model is at F16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 5.6 ms Runs great
BART Large MNLIMeta ยท 410M 56Good quality (B) ยท 56 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. 9.5 s Very slow 56Good quality (B) ยท 56 of 100: how capable the model is at F16, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 6.2 ms Runs great
DynaGuard 1.7BUniversity of Maryland ยท 2.0B 69Good quality (B) ยท 69 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. 193.9 s Very slowtight fit 66Good quality (B) ยท 66 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 284 ms Runs well
ShieldGemma 2BGoogle ยท 2.6B 55Good quality (B) ยท 55 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. 112.0 s Very slowtight fit 58Good quality (B) ยท 58 of 100: how capable the model is at F16, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. 35 ms Runs great
Llama Guard 3 1BMeta ยท 1.5B 55Good quality (B) ยท 55 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. 67.0 s Very slow 56Good quality (B) ยท 56 of 100: how capable the model is at F16, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 48 ms Runs great
Shieldstral 1.0 3BMistral AI ยท 3.9B 81Great quality (A) ยท 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. 165.3 s Very slowfar over memory limit 86Excellent quality (A+) ยท 86 of 100: how capable the model is at F16, whatever it runs on. Match score 93/100 (quality, fit, speed and memory use) sets the ranking. 50 ms Runs great
NuExtract 2.0 4BNuMind ยท 3.8B 72Great quality (A) ยท 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking. 104.3 s Very slowfar over memory limit 73Great quality (A) ยท 73 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 108 ms Runs great
Skywork Reward V2 Llama 3.1 8BSkywork ยท 7.5B 90Excellent quality (A+) ยท 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 12/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 90Excellent quality (A+) ยท 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 144 ms Runs well
GLiDE 27BFastino ยท 26B 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 86Excellent quality (A+) ยท 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
Qwen3 Reranker 4BAlibaba ยท 4.0B 85Excellent quality (A+) ยท 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 90Excellent quality (A+) ยท 90 of 100: how capable the model is at F16, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking. 52 ms Runs great
Qwen3 Reranker 8BAlibaba ยท 8.2B 85Excellent quality (A+) ยท 85 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 90Excellent quality (A+) ยท 90 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking. 88 ms Runs great
Skywork Reward V2 Qwen3 8BSkywork ยท 7.6B 84Great quality (A) ยท 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. 146 ms Runs well
zerank-2ZeroEntropy ยท 4.0B 83Great quality (A) ยท 83 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 88Excellent quality (A+) ยท 88 of 100: how capable the model is at F16, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking. 52 ms Runs great
CompassJudger-2 32BOpenCompass ยท 33B 81Great quality (A) ยท 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 81Great quality (A) ยท 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. 4.5 s Very slowtight fit
Qwen3Guard Gen 8BAlibaba ยท 8.2B 81Great quality (A) ยท 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 11/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 85Excellent quality (A+) ยท 85 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 88/100 (quality, fit, speed and memory use) sets the ranking. 132 ms Runs great
gpt-oss-safeguard 120BOpenAI ยท 117B (5.1B active) 80Great quality (A) ยท 80 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run
LemonDecider 4BAnt Group (inclusionAI) ยท 4.5B 80Great quality (A) ยท 80 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 80Great quality (A) ยท 80 of 100: how capable the model is at F16, whatever it runs on. Match score 91/100 (quality, fit, speed and memory use) sets the ranking. 28 ms Runs great
Granite Guardian 4.1 8BIBM ยท 8.4B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 135 ms Runs well
Intern-Decision 4BShanghai AI Lab ยท 4.5B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at F16, whatever it runs on. Match score 92/100 (quality, fit, speed and memory use) sets the ranking. 39 ms Runs great
Llama 3.1 Nemotron Safety Guard 8B v3NVIDIA ยท 8.0B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 83Great quality (A) ยท 83 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. 129 ms Runs great
Qwen3Guard Gen 4BAlibaba ยท 4.4B 79Great quality (A) ยท 79 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 84Great quality (A) ยท 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. 84 ms Runs great
gpt-oss-safeguard 20BOpenAI ยท 22B (3.6B active) 76Great quality (A) ยท 76 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 76Great quality (A) ยท 76 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 1.0 s Slow
Granite Guardian 3.3 8BIBM ยท 8.2B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 79Great quality (A) ยท 79 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. 131 ms Runs great
Selene 1 Mini 8BAtla ยท 8.0B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 77Great quality (A) ยท 77 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 1.3 s Slow
WildGuard 7BAi2 ยท 7.3B 76Great quality (A) ยท 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 79Great quality (A) ยท 79 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 86/100 (quality, fit, speed and memory use) sets the ranking. 117 ms Runs great
DynaGuard 8BUniversity of Maryland ยท 8.2B 75Great quality (A) ยท 75 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 75Great quality (A) ยท 75 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. 1.1 s Slow
CompassJudger-2 7BOpenCompass ยท 7.6B 74Great quality (A) ยท 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 75Great quality (A) ยท 75 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking. 1.3 s Slow
Nemotron Content Safety Reasoning 4BNVIDIA ยท 4.3B 74Great quality (A) ยท 74 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 10/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 77Great quality (A) ยท 77 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. 842 ms Slow
Prometheus 2 8x7BPrometheus Eval ยท 47B (13B active) 70Great quality (A) ยท 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 9/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 70Great quality (A) ยท 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. 7.4 s Very slowoffloads to RAM
Llama Guard 3 8BMeta ยท 8.0B 64Good quality (B) ยท 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 67Good quality (B) ยท 67 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. 129 ms Runs great
Prometheus 2 7BPrometheus Eval ยท 7.2B 64Good quality (B) ยท 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 1.2 s Slow
GLIDERPatronus AI ยท 3.8B 62Good quality (B) ยท 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 737 ms Slow
Llama Guard 4 12BMeta ยท 12B 62Good quality (B) ยท 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 65Good quality (B) ยท 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 191 ms Runs well
ShieldGemma 27BGoogle ยท 27B 62Good quality (B) ยท 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 62Good quality (B) ยท 62 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 77/100 (quality, fit, speed and memory use) sets the ranking. 260 ms Runs well
Flow Judge v0.1Flow AI ยท 3.8B 60Good quality (B) ยท 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 63Good quality (B) ยท 63 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. 737 ms Slow
ShieldGemma 9BGoogle ยท 9.2B 60Good quality (B) ยท 60 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 8/100 (quality, fit, speed and memory use) sets the ranking. โ€” Won't run 64Good quality (B) ยท 64 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. 98 ms Runs great

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked โ€œMeasuredโ€ come from a linked benchmark.

Was this page helpful?