Apple M5 Ultra (256 GB) spec sheet
What the Apple M5 Ultra (256 GB) can run
We check your GPU or device against every model. Browsers can’t see how much memory you have, so confirm your exact model below.
- Unified memory
- 256 GB
- Usable for models
- 192 GB
- Memory bandwidth
- 1,200 GB/s
- FP16 compute
- 37.7 TFLOPS
Open in Model Finder Best LLMs by task Compare with another GPU
LLMs on the Apple M5 Ultra (256 GB)
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 93Excellent quality (A+) · 93 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. | Nemotron 3.5 Super VL 120B-A12B | Q8_0 | 148 GB | 44.7 tok/s | Runs great |
| 93Excellent quality (A+) · 93 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 97/100 (quality, fit, speed and memory use) sets the ranking. | Qwen3.8 Flash Next (125B-A6B) | Q5_K_M | 144 GB | 108 tok/s | Runs great |
| 92Excellent quality (A+) · 92 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. | Ornith 1.5 122B-A10B | Q8_0 | 149 GB | 51.8 tok/s | Runs great |
| 92Excellent quality (A+) · 92 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. | Qwen3.5 122B-A10B | Q8_0 | 149 GB | 51.8 tok/s | Runs great |
| 90Excellent quality (A+) · 90 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking. | Step 3.7 Flash (198B-A11B) | Q5_K_M | 159 GB | 71.4 tok/s | Runs great |
| 89Excellent quality (A+) · 89 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 95/100 (quality, fit, speed and memory use) sets the ranking. | Step 3.5 Flash (196B-A11B) | Q5_K_M | 158 GB | 71.4 tok/s | Runs great |
| 88Excellent quality (A+) · 88 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. | MiniMax M2.7 (230B-A10B) | Q4_K_M | 158 GB | 88.9 tok/s | Runs great |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking. | Nemotron 3 Super 120B-A12B | Q8_0 | 147 GB | 44.7 tok/s | Runs great |
| 87Excellent quality (A+) · 87 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. | MiniMax M2.5 (230B-A10B) | Q4_K_M | 158 GB | 88.9 tok/s | Runs great |
| 89Excellent quality (A+) · 89 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. | Inkling Small (276B-A12B) | Q4_K_M | 178 GB | 78.0 tok/s | Runs welltight fit |
| 86Excellent quality (A+) · 86 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. | MiniMax M2.1 (230B-A10B) | Q4_K_M | 158 GB | 88.9 tok/s | Runs great |
| 86Excellent quality (A+) · 86 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. | Qwen3 235B-A22B Thinking 2507 | Q4_K_M | 161 GB | 48.3 tok/s | Runs great |
| 86Excellent quality (A+) · 86 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. | Laguna S 2.1 (118B-A8B) | Q8_0 | 142 GB | 61.8 tok/s | Runs great |
| 85Excellent quality (A+) · 85 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. | Qwen3 235B-A22B Instruct 2507 | Q4_K_M | 161 GB | 48.3 tok/s | Runs great |
| 92Excellent quality (A+) · 92 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. | Nemotron 3.5 Bolt 74B-A9B | Q8_0 | 95.4 GB | 56.4 tok/s | Runs great |
| 84Great quality (A) · 84 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. | MiniMax M2 (230B-A10B) | Q4_K_M | 158 GB | 88.9 tok/s | Runs great |
| 84Great quality (A) · 84 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. | Qwen3-VL 235B-A22B Instruct | Q4_K_M | 161 GB | 48.3 tok/s | Runs great |
| 91Excellent quality (A+) · 91 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. | Kolibri 1 (78B-A3.5B) | Q8_0 | 98.9 GB | 109 tok/s | Runs great |
| 93Excellent quality (A+) · 93 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. | Qwen3.6 35B-A3B | F16 | 88.0 GB | 69.2 tok/s | Runs great |
| 83Great quality (A) · 83 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. | Command A+ (218B-A25B) | Q4_K_M | 150 GB | 43.3 tok/s | Runs great |
| 83Great quality (A) · 83 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. | K-EXAONE 236B-A23B | Q4_K_M | 161 GB | 46.5 tok/s | Runs great |
| 83Great quality (A) · 83 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. | Qwen3 235B-A22B | Q4_K_M | 161 GB | 48.3 tok/s | Runs great |
| 85Excellent quality (A+) · 85 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. | Solar Open 2 250B | Q4_K_M | 170 GB | 65.8 tok/s | Runs welltight fit |
| 93Excellent quality (A+) · 93 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. | Qwen3.5 35B-A3B | F16 | 88.0 GB | 69.2 tok/s | Runs great |
| 93Excellent quality (A+) · 93 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. | MiniCPM-V 4.7 35B-A3B | F16 | 86.5 GB | 69.2 tok/s | Runs great |
| 83Great quality (A) · 83 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. | GLM-4.6V | Q8_0 | 132 GB | 44.7 tok/s | Runs great |
| 90Excellent quality (A+) · 90 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. | Flint (40B-A2.5B) | F16 | 95.8 GB | 78.3 tok/s | Runs great |
| 82Great quality (A) · 82 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. | Ling 3.0 Flash (127B-A5B) | Q8_0 | 151 GB | 85.5 tok/s | Runs great |
| 92Excellent quality (A+) · 92 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. | Nex N2 Mini (35B-A3B) | F16 | 86.3 GB | 69.2 tok/s | Runs great |
| 91Excellent quality (A+) · 91 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. | Sage-A1.5 35B-A3B | F16 | 88.0 GB | 69.2 tok/s | Runs great |
| 80Great quality (A) · 80 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. | Mistral Small 4 (119B-A6B) | Q8_0 | 143 GB | 76.4 tok/s | Runs great |
| 91Excellent quality (A+) · 91 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. | KAT-Coder V2.5 Dev (35B-A3B) | F16 | 85.4 GB | 69.2 tok/s | Runs great |
| 90Excellent quality (A+) · 90 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. | Ornith 1.5 35B-A3B | F16 | 88.0 GB | 69.2 tok/s | Runs great |
| 86Excellent quality (A+) · 86 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. | Qwen3-Next 80B-A3B Thinking | Q8_0 | 102 GB | 118 tok/s | Runs great |
| 80Great quality (A) · 80 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. | GLM-4.5V | Q8_0 | 132 GB | 44.7 tok/s | Runs great |
| 80Great quality (A) · 80 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. | Ling 2.6 Flash (104B-A7.4B) | Q8_0 | 130 GB | 65.5 tok/s | Runs great |
| 80Great quality (A) · 80 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. | Solar Open 100B | Q8_0 | 126 GB | 44.7 tok/s | Runs great |
| 79Great quality (A) · 79 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. | Baichuan M3 235B-A22B | Q4_K_M | 161 GB | 48.3 tok/s | Runs great |
| 86Excellent quality (A+) · 86 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. | Qwen3 Coder Next (80B-A3B) | Q8_0 | 101 GB | 118 tok/s | Runs great |
| 81Great quality (A) · 81 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | Nex N2.6 (321B-A18B) | Q3_K_M | 173 GB | 70.3 tok/s | Runs welltight fit |
| 81Great quality (A) · 81 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. | InternVL3.5 241B-A28B | Q4_K_M | 164 GB | 39.3 tok/s | Runs welltight fit |
| 89Excellent quality (A+) · 89 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. | Apodex 1.1 Mini (35B-A3B) | F16 | 88.0 GB | 69.2 tok/s | Runs great |
| 89Excellent quality (A+) · 89 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. | Ornith 1.0 35B-A3B | F16 | 88.0 GB | 69.2 tok/s | Runs great |
| 81Great quality (A) · 81 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | GLM-5.3 Flash (320B-A18B) | Q3_K_M | 173 GB | 70.3 tok/s | Runs welltight fit |
| 81Great quality (A) · 81 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 90/100 (quality, fit, speed and memory use) sets the ranking. | MiMo V2.6 Flash (310B-A15B) | Q3_K_M | 168 GB | 80.7 tok/s | Runs welltight fit |
| 84Great quality (A) · 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. | Qwen3 Next 80B-A3B Instruct | Q8_0 | 102 GB | 118 tok/s | Runs great |
| 78Great quality (A) · 78 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. | GLM-4.5 Air (106B-A12B) | Q8_0 | 135 GB | 44.7 tok/s | Runs great |
| 78Great quality (A) · 78 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. | Ring Flash 2.0 (100B-A6B) | Q8_0 | 126 GB | 75.5 tok/s | Runs great |
| 79Great quality (A) · 79 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. | Naive N0.5 Flash (309B-A15.5B) | Q3_K_M | 168 GB | 78.8 tok/s | Runs welltight fit |
| 92Excellent quality (A+) · 92 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. | Gemma 4 26B-A4B | F16 | 68.0 GB | 58.3 tok/s | Runs great |
| 77Great quality (A) · 77 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. | DeepSeek V4 Flash (April 2026) | Q3_K_M | 159 GB | 89.5 tok/s | Runs great |
| 79Great quality (A) · 79 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. | Hunyuan Hy3 (295B-A21B) | Q3_K_M | 165 GB | 62.3 tok/s | Runs welltight fit |
| 78Great quality (A) · 78 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. | DeepSeek V4 Flash (284B-A13B) | Q3_K_M | 165 GB | 89.5 tok/s | Runs welltight fit |
| 78Great quality (A) · 78 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. | MiMo V2.5 (310B-A15B) | Q3_K_M | 168 GB | 80.7 tok/s | Runs welltight fit |
| 87Excellent quality (A+) · 87 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. | Agents-A1 35B-A3B | F16 | 86.3 GB | 69.2 tok/s | Runs great |
| 87Excellent quality (A+) · 87 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. | Qwen AgentWorld 35B-A3B | F16 | 85.4 GB | 69.2 tok/s | Runs great |
| 76Great quality (A) · 76 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. | Ling Flash 2.0 (100B-A6B) | Q8_0 | 126 GB | 75.5 tok/s | Runs great |
| 88Excellent quality (A+) · 88 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. | GLM-4.7 Flash (30B-A3B) | F16 | 78.8 GB | 69.2 tok/s | Runs great |
| 75Great quality (A) · 75 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. | dots.llm1 Instruct | Q6_K | 141 GB | 51.2 tok/s | Runs great |
| 75Great quality (A) · 75 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. | Sarvam 105B | Q8_0 | 129 GB | 51.8 tok/s | Runs great |
| 88Excellent quality (A+) · 88 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. | Raon-1 mini Preview (30B-A4B) | F16 | 73.6 GB | 53.1 tok/s | Runs great |
| 76Great quality (A) · 76 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. | MiMo V2 Flash (309B-A15B) | Q3_K_M | 168 GB | 80.7 tok/s | Runs welltight fit |
| 74Great quality (A) · 74 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. | Leanstral 119B-A6B | Q8_0 | 143 GB | 72.1 tok/s | Runs great |
| 84Great quality (A) · 84 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 87/100 (quality, fit, speed and memory use) sets the ranking. | gpt-oss 120B | MXFP4 | 82.9 GB | 140 tok/s | Runs great |
| 85Excellent quality (A+) · 85 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. | Hunyuan Hy-Memory 30B-A3B | F16 | 76.9 GB | 69.2 tok/s | Runs great |
| 74Great quality (A) · 74 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. | Kimi Linear 48B-A3B | F16 | 114 GB | 69.2 tok/s | Runs great |
| 82Great quality (A) · 82 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. | Laguna XS 2.1 (33B-A3B) | F16 | 84.2 GB | 69.2 tok/s | Runs great |
| 70Great quality (A) · 70 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. | DeepSeek Coder V2 (236B) | Q4_K_M | 160 GB | 50.2 tok/s | Runs great |
| 70Great quality (A) · 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. | Groot 65B A6B 261023 | Q8_0 | 142 GB | 76.4 tok/s | Runs great |
| 68Good quality (B) · 68 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. | DeepSeek Coder V2 Instruct (0614) | Q4_K_M | 160 GB | 50.2 tok/s | Runs great |
| 68Good quality (B) · 68 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. | DeepSeek V2.5 1210 | Q4_K_M | 160 GB | 50.2 tok/s | Runs great |
| 92Excellent quality (A+) · 92 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. | Qwen3.8 27B | Q5_K_M | 36.2 GB | 37.7 tok/s | Runs great |
| 68Good quality (B) · 68 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. | Ling 3.0 flash Sante | Q8_0 | 151 GB | 85.5 tok/s | Runs great |
| 74Great quality (A) · 74 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. | Hunyuan A13B (80B-A13B) | Q8_0 | 102 GB | 41.8 tok/s | Runs great |
| 80Great quality (A) · 80 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. | Nemotron Cascade 2 30B-A3B | F16 | 79.2 GB | 62.0 tok/s | Runs great |
| 80Great quality (A) · 80 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. | North Mini Code 1.0 (30B-A3B) | F16 | 77.2 GB | 69.2 tok/s | Runs great |
| 90Excellent quality (A+) · 90 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. | Nanbeige4.3 33B | Q4_K_M | 37.9 GB | 38.3 tok/s | Runs great |
| 91Excellent quality (A+) · 91 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. | Qwen3.6 27B | Q5_K_M | 36.2 GB | 37.7 tok/s | Runs great |
| 66Good quality (B) · 66 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. | DeepSeek V2.5 | Q4_K_M | 160 GB | 50.2 tok/s | Runs great |
| 71Great quality (A) · 71 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. | Llama 4 Scout 17B-16E | Q6_K | 107 GB | 43.6 tok/s | Runs great |
| 67Good quality (B) · 67 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | MiniMax M3.1 Flash (428B-A23B) | Q2_K | 189 GB | 70.6 tok/s | Runs welltight fit |
| 90Excellent quality (A+) · 90 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. | Qwen3.5 27B | Q5_K_M | 36.2 GB | 37.7 tok/s | Runs great |
| 78Great quality (A) · 78 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. | Qwen3 30B-A3B Thinking 2507 | F16 | 77.8 GB | 64.7 tok/s | Runs great |
| 78Great quality (A) · 78 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. | Qwen3 Coder 30B-A3B | F16 | 77.8 GB | 64.7 tok/s | Runs great |
| 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | GLM-4.7 (355B-A32B) | Q2_K | 162 GB | 54.1 tok/s | Runs great |
| 77Great quality (A) · 77 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. | Nemotron 3 Nano 30B-A3B | F16 | 79.2 GB | 62.0 tok/s | Runs great |
| 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 84/100 (quality, fit, speed and memory use) sets the ranking. | Intern-S2 (397B-A17B) | Q2_K | 178 GB | 88.6 tok/s | Runs welltight fit |
| 77Great quality (A) · 77 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. | Qwen3 30B-A3B Instruct 2507 | F16 | 77.8 GB | 64.7 tok/s | Runs great |
| 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | Nex N2 Pro (397B-A17B) | Q2_K | 175 GB | 88.6 tok/s | Runs welltight fit |
| 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | Qwen3.5 397B-A17B | Q2_K | 178 GB | 88.6 tok/s | Runs welltight fit |
| 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | MiniMax M3 (427B) | Q2_K | 188 GB | 66.1 tok/s | Runs welltight fit |
| 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | K2-Horizon (375B-A23B) | Q2_K | 170 GB | 70.6 tok/s | Runs welltight fit |
| 76Great quality (A) · 76 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. | Qwen3-VL 30B-A3B Thinking | F16 | 78.9 GB | 64.7 tok/s | Runs great |
| 65Good quality (B) · 65 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | Ornith 1.5 397B-A17B | Q2_K | 178 GB | 88.6 tok/s | Runs welltight fit |
| 87Excellent quality (A+) · 87 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. | Muse Glimmer 30B | Q4_K_M | 34.2 GB | 42.6 tok/s | Runs great |
| 86Excellent quality (A+) · 86 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. | Gemma 4 31B | Q4_K_M | 36.6 GB | 40.6 tok/s | Runs great |
| 87Excellent quality (A+) · 87 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. | TMax 27B | Q5_K_M | 35.9 GB | 38.3 tok/s | Runs great |
| 78Great quality (A) · 78 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. | DiffusionGemma 26B-A4B IT | F16 | 68.0 GB | 56.1 tok/s | Runs great |
| 74Great quality (A) · 74 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. | Qwen3-VL 30B-A3B Instruct | F16 | 78.9 GB | 64.7 tok/s | Runs great |
| 74Great quality (A) · 74 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. | Qwen3 30B-A3B | F16 | 77.8 GB | 64.7 tok/s | Runs great |
| 67Good quality (B) · 67 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 82/100 (quality, fit, speed and memory use) sets the ranking. | ERNIE 4.5 300B-A47B | Q3_K_M | 165 GB | 31.3 tok/s | Runs welltight fit |
| 85Excellent quality (A+) · 85 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. | Fara 1.5 27B | Q5_K_M | 35.9 GB | 38.3 tok/s | Runs great |
| 70Great quality (A) · 70 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. | Qwen3-Omni 30B-A3B Instruct | F16 | 87.3 GB | 64.7 tok/s | Runs great |
| 72Great quality (A) · 72 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. | Nemotron 3.5 Lightning 30B-A3B | F16 | 79.2 GB | 69.2 tok/s | Runs great |
| 83Great quality (A) · 83 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | UI-Mate 27B | Q5_K_M | 35.9 GB | 38.3 tok/s | Runs great |
| 58Good quality (B) · 58 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. | DeepSeek V2 Chat | Q4_K_M | 160 GB | 50.2 tok/s | Runs great |
| 58Good quality (B) · 58 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 81/100 (quality, fit, speed and memory use) sets the ranking. | GLM-4.6 (355B-A32B) | Q2_K | 162 GB | 54.1 tok/s | Runs great |
| 70Great quality (A) · 70 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. | Sarvam 30B | F16 | 80.4 GB | 80.4 tok/s | Runs great |
| 81Great quality (A) · 81 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. | Llama 3.1 Nemotron Ultra 253B v1 | Q4_K_M | 175 GB | 5.0 tok/s | Slowtight fit |
| 74Great quality (A) · 74 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. | LFM2.5 24B-A2B | F16 | 64.1 GB | 79.3 tok/s | Runs great |
| 83Great quality (A) · 83 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. | Gemma 4 12B | Q8_0 | 29.3 GB | 54.0 tok/s | Runs great |
| 57Good quality (B) · 57 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | GLM-4.5 (355B-A32B) | Q2_K | 162 GB | 54.1 tok/s | Runs great |
| 81Great quality (A) · 81 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking. | Devstral 2 (123B) | Q4_K_M | 95.2 GB | 10.1 tok/s | Runs well |
| 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | Trinity Large Thinking (400B-A13B) | Q2_K | 176 GB | 107 tok/s | Runs welltight fit |
| 80Great quality (A) · 80 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. | K2 Horizon 7B | Q8_0 | 26.7 GB | 71.7 tok/s | Runs great |
| 77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | EXAONE 4.5 33B | Q4_K_M | 37.7 GB | 36.9 tok/s | Runs great |
| 70Great quality (A) · 70 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. | ERNIE 4.5 21B-A3B Thinking | F16 | 60.1 GB | 69.2 tok/s | Runs great |
| 54Fair quality (C) · 54 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | NorthCUA Pro | Q2_K | 175 GB | 88.6 tok/s | Runs welltight fit |
| 70Great quality (A) · 70 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. | gpt-oss 20B | F16 | 58.0 GB | 60.7 tok/s | Runs great |
| 78Great quality (A) · 78 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. | SenseNova Loop 9B Preview | Q8_0 | 26.7 GB | 66.8 tok/s | Runs great |
| 70Great quality (A) · 70 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. | Maple Preview (20B-A1B) | F16 | 56.5 GB | 107 tok/s | Runs great |
| 62Good quality (B) · 62 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. | Axion 1.0 35B | F16 | 86.3 GB | 69.2 tok/s | Runs great |
| 78Great quality (A) · 78 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. | MiMo V2.6 Distill Qwen 9B | Q8_0 | 26.2 GB | 68.6 tok/s | Runs great |
| 77Great quality (A) · 77 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | Command A Reasoning (111B) | Q4_K_M | 84.4 GB | 11.4 tok/s | Runs well |
| 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | Mistral Large 2.1 (2411) | Q4_K_M | 93.7 GB | 10.3 tok/s | Runs well |
| 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking. | Seed-OSS 36B Instruct | Q4_K_M | 40.2 GB | 35.1 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Trinity Mini (26B-A3B) | F16 | 68.4 GB | 69.2 tok/s | Runs great |
| 77Great quality (A) · 77 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. | Ornith 1.5 9B | Q8_0 | 26.5 GB | 66.9 tok/s | Runs great |
| 74Great quality (A) · 74 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. | QwQ 32B | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 75Great quality (A) · 75 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. | Devstral Small 2 (24B) | Q6_K | 37.0 GB | 36.7 tok/s | Runs great |
| 75Great quality (A) · 75 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. | Hermes 4.3 36B | Q4_K_M | 40.2 GB | 35.1 tok/s | Runs great |
| 74Great quality (A) · 74 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. | Qwen3-VL 32B Instruct | Q4_K_M | 38.5 GB | 38.0 tok/s | Runs great |
| 76Great quality (A) · 76 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. | Qwen3.5 9B | Q8_0 | 26.5 GB | 66.9 tok/s | Runs great |
| 76Great quality (A) · 76 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. | Command A (111B) | Q4_K_M | 84.4 GB | 11.4 tok/s | Runs well |
| 74Great quality (A) · 74 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. | Pixtral Large 124B (2411) | Q4_K_M | 94.6 GB | 10.2 tok/s | Runs well |
| 74Great quality (A) · 74 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. | Mistral Large 2 (123B) | Q4_K_M | 93.9 GB | 10.3 tok/s | Runs well |
| 76Great quality (A) · 76 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. | Llama 3.3 Nemotron Super 49B v1.5 | Q4_K_M | 49.1 GB | 25.4 tok/s | Runs well |
| 75Great quality (A) · 75 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. | Ornith 1.0 9B | Q8_0 | 26.5 GB | 66.9 tok/s | Runs great |
| 59Good quality (B) · 59 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. | NorthCUA Mini | F16 | 86.3 GB | 69.2 tok/s | Runs great |
| 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | OpenReasoning Nemotron 32B | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | Skywork OR1 32B | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | Qwen3 32B | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | Granite 4.2 30B | Q4_K_M | 36.0 GB | 43.3 tok/s | Runs great |
| 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | Olmo 3.1 32B Think | Q4_K_M | 36.4 GB | 39.4 tok/s | Runs great |
| 50Fair quality (C) · 50 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | Llama 4 Maverick (17B x 128E) | Q2_K | 178 GB | 88.6 tok/s | Runs welltight fit |
| 76Great quality (A) · 76 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | Kimi Dev 72B | Q4_K_M | 63.0 GB | 17.4 tok/s | Runs well |
| 72Great quality (A) · 72 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. | Apriel 1.5 15B Thinker | Q8_0 | 33.3 GB | 43.5 tok/s | Runs great |
| 74Great quality (A) · 74 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. | TMax 9B | Q8_0 | 25.7 GB | 72.1 tok/s | Runs great |
| 72Great quality (A) · 72 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. | Phi-4 Reasoning Plus (14B) | Q8_0 | 33.2 GB | 44.0 tok/s | Runs great |
| 70Great quality (A) · 70 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. | Mellum2.1 Thinking (12B-A2.5B) | F16 | 40.4 GB | 78.3 tok/s | Runs great |
| 72Great quality (A) · 72 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | InternVL3.5 38B | Q4_K_M | 41.5 GB | 33.0 tok/s | Runs great |
| 70Great quality (A) · 70 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 76/100 (quality, fit, speed and memory use) sets the ranking. | EXAONE Deep 32B | Q4_K_M | 37.6 GB | 39.6 tok/s | Runs great |
| 64Good quality (B) · 64 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. | ERNIE 4.5 21B-A3B | F16 | 60.3 GB | 69.2 tok/s | Runs great |
| 71Great quality (A) · 71 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. | Magistral Small 1.2 (24B) | Q6_K | 37.0 GB | 36.7 tok/s | Runs great |
| 70Great quality (A) · 70 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. | GLM-4 32B 0414 | Q4_K_M | 36.3 GB | 38.9 tok/s | Runs great |
| 70Great quality (A) · 70 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. | GLM-Z1 32B (0414) | Q4_K_M | 36.3 GB | 38.9 tok/s | Runs great |
| 70Great quality (A) · 70 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. | Phi-4 Reasoning Vision 15B | Q8_0 | 33.7 GB | 42.7 tok/s | Runs great |
| 70Great quality (A) · 70 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. | Phi-4 Reasoning (14B) | Q8_0 | 33.2 GB | 44.0 tok/s | Runs great |
| 69Good quality (B) · 69 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. | Gemma 3 27B IT | Q5_K_M | 36.2 GB | 38.2 tok/s | Runs great |
| 70Great quality (A) · 70 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. | Nemotron 3.5 Bolt E4B | F16 | 32.1 GB | 66.1 tok/s | Runs great |
| 68Good quality (B) · 68 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. | DeepSeek R1 Distill Qwen 32B | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 69Good quality (B) · 69 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. | Devstral Small 1.1 (24B) | Q6_K | 36.6 GB | 37.3 tok/s | Runs great |
| 69Good quality (B) · 69 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. | Magistral Small 2507 | Q6_K | 36.6 GB | 37.3 tok/s | Runs great |
| 68Good quality (B) · 68 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. | Granite 4.1 30B | Q4_K_M | 35.7 GB | 43.9 tok/s | Runs great |
| 68Good quality (B) · 68 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. | EXAONE 4.0 32B | Q4_K_M | 36.2 GB | 39.6 tok/s | Runs great |
| 68Good quality (B) · 68 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. | Olmo 3.1 32B Instruct | Q4_K_M | 36.4 GB | 39.4 tok/s | Runs great |
| 70Great quality (A) · 70 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. | UI-Mate 9B | Q8_0 | 26.2 GB | 68.6 tok/s | Runs great |
| 70Great quality (A) · 70 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. | UI-Venus 2 9B | Q8_0 | 26.2 GB | 68.6 tok/s | Runs great |
| 67Good quality (B) · 67 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 74/100 (quality, fit, speed and memory use) sets the ranking. | MedGemma 27B Text IT | Q5_K_M | 35.9 GB | 38.8 tok/s | Runs great |
| 68Good quality (B) · 68 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. | Qwen3 14B | Q8_0 | 33.0 GB | 43.6 tok/s | Runs great |
| 68Good quality (B) · 68 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. | Gemma 4 E4B | F16 | 32.1 GB | 66.1 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Baichuan M2 32B | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 66Good quality (B) · 66 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. | OpenHands LM 32B | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Qwen2.5 32B Instruct | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 67Good quality (B) · 67 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. | Magistral Small 2506 | Q6_K | 36.6 GB | 37.3 tok/s | Runs great |
| 70Great quality (A) · 70 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. | InternVL3 78B | Q4_K_M | 66.5 GB | 16.2 tok/s | Runs well |
| 70Great quality (A) · 70 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. | Llama 3.3 70B Instruct | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 70Great quality (A) · 70 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. | Hermes 4 70B | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 62Good quality (B) · 62 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. | Granite 4.0 H Small (32B-A9B) | Q8_0 | 50.2 GB | 56.4 tok/s | Runs great |
| 65Good quality (B) · 65 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. | Mistral Small 3.2 24B | Q5_K_M | 34.3 GB | 43.6 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Mixtral 8x22B Instruct | Q4_K_M | 104 GB | 29.3 tok/s | Runs well |
| 68Good quality (B) · 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. | Granite 4.2 8B | Q8_0 | 26.6 GB | 73.5 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Apriel Nemotron 15B Thinker | Q8_0 | 33.5 GB | 43.1 tok/s | Runs great |
| 68Good quality (B) · 68 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. | Nanbeige4.2 3B | F16 | 25.8 GB | 71.4 tok/s | Runs great |
| 66Good quality (B) · 66 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. | AceReason Nemotron 14B | Q8_0 | 33.2 GB | 43.7 tok/s | Runs great |
| 66Good quality (B) · 66 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. | OpenReasoning Nemotron 14B | Q8_0 | 33.2 GB | 43.7 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Qwen2.5 Coder 14B Instruct | Q8_0 | 33.2 GB | 43.7 tok/s | Runs great |
| 69Good quality (B) · 69 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. | Qwen2.5-VL 72B Instruct | Q4_K_M | 63.4 GB | 17.3 tok/s | Runs well |
| 69Good quality (B) · 69 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. | Qwen2.5 72B Instruct | Q4_K_M | 63.0 GB | 17.4 tok/s | Runs well |
| 67Good quality (B) · 67 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | Mistral Medium 3.5 (128B) | Q3_K_M | 81.5 GB | 12.8 tok/s | Runs well |
| 66Good quality (B) · 66 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. | Hermes 4 14B | Q8_0 | 33.0 GB | 43.7 tok/s | Runs great |
| 65Good quality (B) · 65 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. | Terminal RSI 27B | Q5_K_M | 36.2 GB | 37.7 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Ministral 3 14B Reasoning | Q8_0 | 32.1 GB | 46.3 tok/s | Runs great |
| 66Good quality (B) · 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. | InternVL3 38B | Q4_K_M | 41.5 GB | 33.0 tok/s | Runs great |
| 68Good quality (B) · 68 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. | Nemotron Labs Dory 1.0 4B | F16 | 24.4 GB | 74.8 tok/s | Runs great |
| 65Good quality (B) · 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | Reka Flash 3.1 (21B) | Q6_K | 34.2 GB | 42.1 tok/s | Runs great |
| 64Good quality (B) · 64 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. | Hemmingway 1.5 | Q5_K_M | 35.6 GB | 39.0 tok/s | Runs great |
| 64Good quality (B) · 64 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. | Qwen2.5-VL 32B Instruct | Q4_K_M | 38.5 GB | 37.9 tok/s | Runs great |
| 65Good quality (B) · 65 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. | Devstral Small 2505 | Q6_K | 36.6 GB | 37.3 tok/s | Runs great |
| 68Good quality (B) · 68 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. | DeepSeek R1 Distill Llama 70B | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 68Good quality (B) · 68 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. | Tülu 3 70B | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 66Good quality (B) · 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. | Command R+ (08-2024) 104B | Q4_K_M | 81.4 GB | 12.2 tok/s | Runs well |
| 63Good quality (B) · 63 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. | Qwen2.5 Coder 32B Instruct | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 64Good quality (B) · 64 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. | Mistral Small 3.1 24B | Q5_K_M | 34.3 GB | 43.7 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Kimi-VL A3B Thinking 2506 | F16 | 49.0 GB | 69.2 tok/s | Runs great |
| 56Good quality (B) · 56 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. | LFM2 24B-A2B | F16 | 63.8 GB | 82.6 tok/s | Runs great |
| 64Good quality (B) · 64 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. | InternVL3 14B | Q8_0 | 33.6 GB | 42.7 tok/s | Runs great |
| 64Good quality (B) · 64 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. | SenseNova 6.8 Code2Video | Q5_K_M | 35.9 GB | 38.3 tok/s | Runs great |
| 64Good quality (B) · 64 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. | DeepCoder 14B Preview | Q8_0 | 33.2 GB | 43.7 tok/s | Runs great |
| 64Good quality (B) · 64 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. | Phi-4 (14B) | Q8_0 | 33.2 GB | 43.9 tok/s | Runs great |
| 64Good quality (B) · 64 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. | Ministral 3 14B | Q8_0 | 32.1 GB | 46.3 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Spark-X2.6 4B | F16 | 24.5 GB | 72.4 tok/s | Runs great |
| 56Good quality (B) · 56 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. | Phi-3.5 MoE (42B-A6.6B) | Q8_0 | 61.4 GB | 71.3 tok/s | Runs great |
| 62Good quality (B) · 62 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. | QwQ 32B Preview | Q4_K_M | 38.1 GB | 38.7 tok/s | Runs great |
| 63Good quality (B) · 63 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. | Mistral Small 3 24B (2501) | Q6_K | 36.6 GB | 37.3 tok/s | Runs great |
| 66Good quality (B) · 66 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. | Llama 3.1 70B Instruct | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 57Good quality (B) · 57 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. | Groot 19B A3B 261023 | F16 | 55.0 GB | 69.2 tok/s | Runs great |
| 62Good quality (B) · 62 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. | DeepSeek R1 Distill Qwen 14B | Q8_0 | 33.2 GB | 43.7 tok/s | Runs great |
| 62Good quality (B) · 62 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. | Qwen2.5 14B Instruct | Q8_0 | 33.2 GB | 43.9 tok/s | Runs great |
| 61Good quality (B) · 61 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. | Codestral 22B | Q6_K | 36.1 GB | 39.6 tok/s | Runs great |
| 62Good quality (B) · 62 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. | JanusCoder 14B | Q8_0 | 33.0 GB | 43.7 tok/s | Runs great |
| 49Fair quality (C) · 49 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | Jamba 1.5 Large | Q2_K | 176 GB | 20.7 tok/s | Runs welltight fit |
| 50Fair quality (C) · 50 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. | Qwen2 57B-A14B Instruct | Q8_0 | 77.3 GB | 39.2 tok/s | Runs great |
| 64Good quality (B) · 64 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. | QVQ 72B Preview | Q4_K_M | 63.4 GB | 17.3 tok/s | Runs well |
| 64Good quality (B) · 64 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. | Hermes 3 (Llama 3.1 70B) | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 60Good quality (B) · 60 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. | Falcon-H1 34B Instruct | Q4_K_M | 37.7 GB | 37.7 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Nous Hermes 2 Mixtral 8x7B DPO | Q8_0 | 66.5 GB | 42.0 tok/s | Runs great |
| 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | Gemma 2 27B IT | Q5_K_M | 37.0 GB | 38.5 tok/s | Runs great |
| 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | Voxtral Small 24B | Q5_K_M | 34.5 GB | 43.2 tok/s | Runs great |
| 60Good quality (B) · 60 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | Dolphin Mistral 24B Venice Edition | Q5_K_M | 34.3 GB | 43.6 tok/s | Runs great |
| 62Good quality (B) · 62 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. | Qwen3-VL 8B Instruct | Q8_0 | 26.5 GB | 73.6 tok/s | Runs great |
| 62Good quality (B) · 62 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. | Gemma 4 E2B | F16 | 26.3 GB | 129 tok/s | Runs great |
| 62Good quality (B) · 62 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. | InternVL3.5 8B | Q8_0 | 26.2 GB | 75.7 tok/s | Runs great |
| 62Good quality (B) · 62 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. | DeepSeek R1 0528 Qwen3 8B | Q8_0 | 25.8 GB | 78.8 tok/s | Runs great |
| 62Good quality (B) · 62 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. | Llama 3.2 90B Vision Instruct | Q4_K_M | 73.4 GB | 14.3 tok/s | Runs well |
| 50Fair quality (C) · 50 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. | Jamba 1.5 Mini | Q8_0 | 70.8 GB | 44.7 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Qwen2.5 14B Instruct 1M | Q8_0 | 33.2 GB | 43.7 tok/s | Runs great |
| 62Good quality (B) · 62 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. | Qwen3.5 4B | F16 | 25.5 GB | 63.9 tok/s | Runs great |
| 59Good quality (B) · 59 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | Solar Pro Preview 22B | Q6_K | 35.5 GB | 39.8 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Ling 3.0 Tiny (8B-A1.3B) | F16 | 32.0 GB | 114 tok/s | Runs great |
| 58Good quality (B) · 58 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. | Command R (08-2024) 32B | Q4_K_M | 37.0 GB | 39.3 tok/s | Runs great |
| 62Good quality (B) · 62 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. | Apertus v1.5 70B | Q4_K_M | 62.6 GB | 17.6 tok/s | Runs well |
| 62Good quality (B) · 62 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. | Llama 3.1 Nemotron 70B Instruct | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 59Good quality (B) · 59 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | Dolphin 3.0 R1 Mistral 24B | Q6_K | 36.6 GB | 37.3 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Gemma 3 12B IT | Q8_0 | 29.6 GB | 52.9 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Mixtral 8x7B Instruct | Q8_0 | 66.5 GB | 42.0 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Command R+ 104B (04-2024) | Q4_K_M | 81.4 GB | 12.2 tok/s | Runs well |
| 61Good quality (B) · 61 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. | TMax 4B | F16 | 24.6 GB | 70.7 tok/s | Runs great |
| 60Good quality (B) · 60 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. | GLM-4.1V 9B Thinking | Q8_0 | 27.2 GB | 62.8 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Mage-VL-Cowork 4.7B | F16 | 26.6 GB | 62.8 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Granite 4.1 8B | Q8_0 | 26.6 GB | 73.5 tok/s | Runs great |
| 60Good quality (B) · 60 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. | MiniCPM-V 4.5 8B | Q8_0 | 26.4 GB | 74.2 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Qwen3 8B | Q8_0 | 25.8 GB | 78.8 tok/s | Runs great |
| 57Good quality (B) · 57 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | LLM4Decompile 22B V2 | Q6_K | 36.1 GB | 39.6 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Nemotron Nano 9B v2 | Q8_0 | 25.5 GB | 72.6 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Agents-A1 4B | F16 | 25.3 GB | 65.6 tok/s | Runs great |
| 57Good quality (B) · 57 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. | Aya Expanse 32B | Q4_K_M | 37.0 GB | 39.3 tok/s | Runs great |
| 57Good quality (B) · 57 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. | OLMo 2 32B Instruct | Q4_K_M | 36.7 GB | 39.4 tok/s | Runs great |
| 60Good quality (B) · 60 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. | Qwen2-VL 72B Instruct | Q4_K_M | 63.4 GB | 17.3 tok/s | Runs well |
| 60Good quality (B) · 60 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. | Qwen2 72B Instruct | Q4_K_M | 63.0 GB | 17.4 tok/s | Runs well |
| 57Good quality (B) · 57 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. | Aya Vision 32B | Q4_K_M | 37.5 GB | 38.3 tok/s | Runs great |
| 56Good quality (B) · 56 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | Hermes 4 405B | Q2_K | 183 GB | 5.1 tok/s | Slowtight fit |
| 60Good quality (B) · 60 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. | MiniCPM5 2B | F16 | 21.4 GB | 118 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Ling Mini 2.0 (16B-A1.4B) | F16 | 48.8 GB | 110 tok/s | Runs great |
| 58Good quality (B) · 58 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. | GLM-Z1 9B (0414) | Q8_0 | 26.3 GB | 68.7 tok/s | Runs great |
| 58Good quality (B) · 58 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. | MiMo-VL 7B RL | Q8_0 | 26.0 GB | 77.7 tok/s | Runs great |
| 58Good quality (B) · 58 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. | Seed-Coder 8B Reasoning | Q8_0 | 25.8 GB | 78.3 tok/s | Runs great |
| 58Good quality (B) · 58 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. | Fara 1.5 4B | F16 | 25.3 GB | 65.6 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | DeepSeek Coder V2 Lite (16B) | F16 | 47.6 GB | 80.4 tok/s | Runs great |
| 63Good quality (B) · 63 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. | Qwen3 Coder 480B-A35B | Q2_K | 210 GB | 42.7 tok/s | Runs wellover GPU memory limit |
| 56Good quality (B) · 56 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. | Mage-VL 4.7B | F16 | 26.6 GB | 62.8 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Kimi-VL A3B Instruct | F16 | 49.0 GB | 69.2 tok/s | Runs great |
| 56Good quality (B) · 56 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. | Ministral 3 8B | Q8_0 | 26.6 GB | 72.4 tok/s | Runs great |
| 53Fair quality (C) · 53 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. | Tülu 3 405B | Q2_K | 183 GB | 5.1 tok/s | Slowtight fit |
| 53Fair quality (C) · 53 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. | Mistral Small 22B (2409) | Q6_K | 36.1 GB | 39.6 tok/s | Runs great |
| 56Good quality (B) · 56 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. | Seed-Coder 8B Instruct | Q8_0 | 25.8 GB | 78.3 tok/s | Runs great |
| 56Good quality (B) · 56 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. | MiMo 7B RL | Q8_0 | 25.5 GB | 82.5 tok/s | Runs great |
| 56Good quality (B) · 56 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. | EXAONE Deep 7.8B | Q8_0 | 25.3 GB | 82.6 tok/s | Runs great |
| 56Good quality (B) · 56 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. | Qwen3 4B Thinking 2507 | F16 | 25.2 GB | 74.0 tok/s | Runs great |
| 56Good quality (B) · 56 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. | InternVL3 8B | Q8_0 | 24.8 GB | 81.3 tok/s | Runs great |
| 54Fair quality (C) · 54 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. | Olmo 3 7B Think | F16 | 32.0 GB | 40.8 tok/s | Runs great |
| 51Fair quality (C) · 51 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | Command R 35B (v01) | Q4_K_M | 48.0 GB | 36.3 tok/s | Runs great |
| 56Good quality (B) · 56 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. | Falcon-H1R 7B | Q8_0 | 24.4 GB | 85.1 tok/s | Runs great |
| 57Good quality (B) · 57 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 68/100 (quality, fit, speed and memory use) sets the ranking. | Apertus 70B Instruct | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 42Fair quality (C) · 42 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | Grok-1 (314B) | Q3_K_M | 172 GB | 18.0 tok/s | Runs welltight fit |
| 55Good quality (B) · 55 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. | Qwen2.5 Coder 7B Instruct | Q8_0 | 24.5 GB | 84.7 tok/s | Runs great |
| 56Good quality (B) · 56 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. | LFM3 1.7B | F16 | 19.5 GB | 172 tok/s | Runs great |
| 54Fair quality (C) · 54 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. | Qwen3-VL 4B Instruct | F16 | 26.0 GB | 67.0 tok/s | Runs great |
| 54Fair quality (C) · 54 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. | Llama 3.1 Nemotron Nano 8B v1 | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | LFM2.5 8B-A1B | F16 | 33.0 GB | 106 tok/s | Runs great |
| 54Fair quality (C) · 54 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. | Qwen3 4B Instruct 2507 | F16 | 25.2 GB | 74.0 tok/s | Runs great |
| 54Fair quality (C) · 54 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. | MiniCPM4 8B | Q8_0 | 24.9 GB | 78.8 tok/s | Runs great |
| 55Good quality (B) · 55 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | Llama 3 70B Instruct | Q4_K_M | 61.7 GB | 18.0 tok/s | Runs well |
| 54Fair quality (C) · 54 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. | OpenReasoning Nemotron 7B | Q8_0 | 24.5 GB | 84.7 tok/s | Runs great |
| 55Good quality (B) · 55 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | Qwen2.5-Math 72B Instruct | Q4_K_M | 61.7 GB | 17.4 tok/s | Runs well |
| 50Fair quality (C) · 50 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. | Llama 3.1 405B Instruct | Q2_K | 183 GB | 5.1 tok/s | Slowtight fit |
| 52Fair quality (C) · 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. | Mistral Nemo 12B Instruct | Q8_0 | 30.2 GB | 52.9 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Falcon 3 10B Instruct | Q8_0 | 28.2 GB | 62.6 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | GLM-4 9B 0414 | Q8_0 | 26.3 GB | 68.7 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | NorthCUA Nano | Q8_0 | 26.2 GB | 68.6 tok/s | Runs great |
| 46Fair quality (C) · 46 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. | Moonlight 16B-A3B Instruct | F16 | 48.1 GB | 69.2 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | MiniCPM-o 2.6 8B | Q8_0 | 25.6 GB | 74.5 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Foundation-Sec 8B Reasoning | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Qwen2.5-VL 7B Instruct | Q8_0 | 25.2 GB | 77.9 tok/s | Runs great |
| 47Fair quality (C) · 47 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | Aya 23 35B | Q4_K_M | 48.0 GB | 36.3 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Hunyuan 7B Instruct | Q8_0 | 25.0 GB | 86.1 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Olmo 3 7B Instruct | F16 | 32.0 GB | 40.8 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Qwen2.5 7B Instruct | Q8_0 | 24.5 GB | 84.7 tok/s | Runs great |
| 49Fair quality (C) · 49 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | InternLM2.5 20B Chat | Q6_K | 33.9 GB | 44.3 tok/s | Runs great |
| 52Fair quality (C) · 52 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. | Granite 4.2 3B | F16 | 24.0 GB | 81.3 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Pixtral 12B (2409) | Q8_0 | 30.4 GB | 52.1 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Gemma 2 9B IT | Q8_0 | 27.4 GB | 69.9 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Qianfan-OCR 4B | F16 | 26.6 GB | 62.8 tok/s | Runs great |
| 49Fair quality (C) · 49 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. | Qwen1.5 110B Chat | Q4_K_M | 86.5 GB | 11.4 tok/s | Runs well |
| 50Fair quality (C) · 50 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. | InternLM3 8B Instruct | Q8_0 | 25.7 GB | 73.4 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | InternVL2.5 8B | Q8_0 | 25.6 GB | 79.9 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | DeepSeek R1 Distill Llama 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Foundation-Sec 1.1 8B Instruct | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Fara 7B | Q8_0 | 25.2 GB | 77.9 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Qwen3 4B | F16 | 25.2 GB | 74.0 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | GLiDE 9B | Q8_0 | 24.6 GB | 81.3 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Qwen2.5 7B Instruct 1M | Q8_0 | 24.5 GB | 84.7 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Falcon-H1 7B Instruct | Q8_0 | 24.4 GB | 85.1 tok/s | Runs great |
| 50Fair quality (C) · 50 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. | Nemotron 3 Nano 4B | F16 | 24.0 GB | 75.0 tok/s | Runs great |
| 47Fair quality (C) · 47 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. | Yi 1.5 34B Chat | Q4_K_M | 37.9 GB | 36.9 tok/s | Runs great |
| 58Good quality (B) · 58 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. | MiniMax-M1 80K | Q2_K | 199 GB | 33.8 tok/s | Runs wellover GPU memory limit |
| 50Fair quality (C) · 50 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. | LFM3-VL 2B | F16 | 19.9 GB | 155 tok/s | Runs great |
| 48Fair quality (C) · 48 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. | Chandra OCR 2 | F16 | 26.8 GB | 56.2 tok/s | Runs great |
| 48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | GLM-4 9B Chat 1M | Q8_0 | 26.7 GB | 68.1 tok/s | Runs great |
| 48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | GLM-4 9B Chat (2024) | Q8_0 | 26.3 GB | 68.7 tok/s | Runs great |
| 48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | Hermes 3 (Llama 3.1 8B) | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | Llama 3.1 8B Instruct | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 48Fair quality (C) · 48 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. | LightOnOCR-3 4B | F16 | 25.3 GB | 65.6 tok/s | Runs great |
| 57Good quality (B) · 57 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. | MiniMax-M1 40K | Q2_K | 199 GB | 33.8 tok/s | Runs wellover GPU memory limit |
| 48Fair quality (C) · 48 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | MiniCPM-V 2.6 8B | Q8_0 | 25.0 GB | 79.7 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | Phi-3 Medium 128K Instruct | Q8_0 | 32.4 GB | 46.3 tok/s | Runs great |
| 46Fair quality (C) · 46 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. | Gemma 3n E4B | F16 | 31.8 GB | 74.4 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | Phi-3 Medium (14B) | Q8_0 | 31.6 GB | 46.3 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 64/100 (quality, fit, speed and memory use) sets the ranking. | Qwen2.5-Omni 7B | Q8_0 | 27.8 GB | 60.2 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Apertus v1.5 8B | Q8_0 | 26.5 GB | 72.6 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | OLMo 2 13B Instruct | Q8_0 | 33.9 GB | 47.1 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | CodeGeeX4 All 9B | Q8_0 | 26.3 GB | 68.7 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Granite 3.2 8B Instruct | Q8_0 | 26.0 GB | 79.0 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Granite 3.3 8B | Q8_0 | 26.0 GB | 79.0 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Ministral 8B (2410) | Q8_0 | 25.7 GB | 80.5 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | DeepHermes 3 Llama 3 8B Preview | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Dolphin 3.0 Llama 3.1 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Tülu 3 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | EXAONE 3.5 7.8B | Q8_0 | 25.3 GB | 82.6 tok/s | Runs great |
| 40Fair quality (C) · 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. | DeepSeek V2 Lite Chat | F16 | 47.6 GB | 80.4 tok/s | Runs great |
| 66Good quality (B) · 66 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Beam 1.0 (501B-A23B) | Q2_K | 217 GB | 59.2 tok/s | Slowfar over GPU memory limit |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Falcon 3 7B Instruct | Q8_0 | 24.8 GB | 86.6 tok/s | Runs great |
| 46Fair quality (C) · 46 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. | Phi-4 Mini Reasoning (3.8B) | F16 | 24.7 GB | 77.5 tok/s | Runs great |
| 46Fair quality (C) · 46 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | DeepSeek R1 Distill Qwen 7B | Q8_0 | 24.5 GB | 84.7 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | Phi-3 Small 8K Instruct | F16 | 31.8 GB | 40.3 tok/s | Runs great |
| 46Fair quality (C) · 46 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. | Granite 4.1 3B | F16 | 23.4 GB | 87.5 tok/s | Runs great |
| 45Fair quality (C) · 45 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. | WeVisOne-VL 4B | F16 | 26.8 GB | 61.6 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | Granite 4.0 H Tiny (7B-A1B) | F16 | 29.9 GB | 129 tok/s | Runs great |
| 45Fair quality (C) · 45 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. | Gemma 3 4B IT | F16 | 24.8 GB | 69.2 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | Phi-4 Multimodal Instruct | F16 | 28.2 GB | 53.4 tok/s | Runs great |
| 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Qwen1.5 32B Chat | Q4_K_M | 37.9 GB | 39.0 tok/s | Runs great |
| 41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | DeepSeek Coder 33B Instruct | Q4_K_M | 38.4 GB | 38.0 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | Llama 3.1 Nemotron Nano 4B v1.1 | F16 | 26.1 GB | 66.0 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Granite 3.1 8B Instruct | Q8_0 | 26.0 GB | 79.0 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | InternVL2 8B | Q8_0 | 25.6 GB | 79.9 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Dolphin 2.9.4 Llama 3.1 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Hermes 2 Pro Llama 3 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | OpenChat 3.6 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Aya Vision 8B | Q8_0 | 25.5 GB | 74.8 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Qwen2-VL 7B Instruct | Q8_0 | 25.2 GB | 77.9 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Command R7B | Q8_0 | 24.8 GB | 80.4 tok/s | Runs great |
| 41Fair quality (C) · 41 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. | InternLM2 Chat 20B | Q6_K | 33.9 GB | 44.3 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | MedGemma 4B IT | F16 | 24.8 GB | 69.2 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | Phi-4 Mini (3.8B) | F16 | 24.7 GB | 77.5 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | Ministral 3 3B | F16 | 24.5 GB | 77.3 tok/s | Runs great |
| 44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Qwen2 7B Instruct | Q8_0 | 24.5 GB | 84.7 tok/s | Runs great |
| 41Fair quality (C) · 41 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. | LLaVA 1.6 34B | Q4_K_M | 38.2 GB | 36.5 tok/s | Runs great |
| 53Fair quality (C) · 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. | MiniMax-Text-01 | Q2_K | 199 GB | 33.8 tok/s | Runs wellover GPU memory limit |
| 44Fair quality (C) · 44 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. | Unlimited OCR 2 (3B-A0.6B) | F16 | 23.1 GB | 156 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | Jamba Reasoning 3B | F16 | 22.4 GB | 93.0 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | LFM2.5-VL 3B | F16 | 22.3 GB | 95.4 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | North Micro Vision | F16 | 21.9 GB | 120 tok/s | Runs great |
| 44Fair quality (C) · 44 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. | LFM2.5 2.6B | F16 | 21.5 GB | 110 tok/s | Runs great |
| 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Llama 3.2 11B Vision Instruct | Q8_0 | 28.6 GB | 60.5 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | DeepSeek Prover V2 7B | F16 | 33.8 GB | 43.1 tok/s | Runs great |
| 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Mistral-NeMo-Minitron 8B Instruct | Q8_0 | 26.2 GB | 76.8 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | StarCoder2 15B Instruct | Q8_0 | 33.5 GB | 40.5 tok/s | Runs great |
| 42Fair quality (C) · 42 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. | Hunyuan 4B Instruct | F16 | 25.6 GB | 70.5 tok/s | Runs great |
| 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | Llama 3 8B Instruct | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 39Basic quality (D) · 39 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | InternLM2 Math Plus 20B | Q6_K | 33.9 GB | 44.3 tok/s | Runs great |
| 42Fair quality (C) · 42 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | Command R7B Arabic | Q8_0 | 24.8 GB | 80.4 tok/s | Runs great |
| 42Fair quality (C) · 42 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. | SmolLM3 3B | F16 | 22.7 GB | 96.6 tok/s | Runs great |
| 42Fair quality (C) · 42 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. | Jamba2 3B | F16 | 22.0 GB | 98.2 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | Nous Hermes 2 SOLAR 10.7B | Q8_0 | 28.1 GB | 60.2 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | SOLAR 10.7B Instruct | Q8_0 | 28.1 GB | 60.2 tok/s | Runs great |
| 37Basic quality (D) · 37 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | CogVLM2 Llama 3 Chat 19B | Q6_K | 36.2 GB | 45.1 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | Phi-3.5 Mini Instruct | F16 | 26.8 GB | 77.9 tok/s | Runs great |
| 42Fair quality (C) · 42 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. | MiniCPM-OCR 1.0 1.3B | F16 | 18.7 GB | 229 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | Yi Coder 9B Chat | Q8_0 | 26.1 GB | 73.1 tok/s | Runs great |
| 41Fair quality (C) · 41 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. | Code Llama 70B Instruct | Q4_K_M | 59.4 GB | 18.4 tok/s | Runs well |
| 40Fair quality (C) · 40 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. | AFM 4.5B | F16 | 25.8 GB | 64.4 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Granite Code 34B Instruct | Q4_K_M | 36.9 GB | 37.6 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | Apertus 8B Instruct | Q8_0 | 25.6 GB | 80.2 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | Aya Expanse 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 42Fair quality (C) · 42 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. | PaddleOCR-VL 1.6 0.9B | F16 | 18.0 GB | 310 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Yi 34B Chat | Q4_K_M | 37.9 GB | 36.9 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | Granite 3.0 8B Instruct | Q8_0 | 25.3 GB | 79.0 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | LFM2 8B-A1B | F16 | 32.7 GB | 106 tok/s | Runs great |
| 40Fair quality (C) · 40 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 61/100 (quality, fit, speed and memory use) sets the ranking. | InternLM2.5 7B Chat | Q8_0 | 25.2 GB | 83.4 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Mistral 7B Instruct v0.3 | F16 | 31.5 GB | 41.0 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Zamba2 7B Instruct | F16 | 31.4 GB | 40.4 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | Granite 4.0 Micro (3B) | F16 | 23.4 GB | 87.5 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | Unlimited OCR (3B-A0.6B) | F16 | 23.1 GB | 156 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Falcon3 Mamba 7B Instruct | F16 | 30.5 GB | 40.9 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | Qwen3.5 2B | F16 | 20.6 GB | 131 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Phi-3.5 Vision Instruct | F16 | 27.5 GB | 71.7 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | TMax 2B | F16 | 19.8 GB | 158 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Code Llama 34B Instruct | Q4_K_M | 38.2 GB | 37.6 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | TeleOCR | F16 | 19.7 GB | 210 tok/s | Runs great |
| 40Fair quality (C) · 40 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. | MiniCPM-V 4.7 1B | F16 | 19.2 GB | 192 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | LLM4Decompile 9B V2 | Q8_0 | 26.1 GB | 73.1 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Dolphin 2.9 Llama 3 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | OLMo 2 7B Instruct | F16 | 32.7 GB | 40.8 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | EXAONE 3.0 7.8B Instruct | Q8_0 | 24.8 GB | 82.6 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | MiniCPM3 4B | F16 | 24.4 GB | 73.1 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Hermes 2 Pro Mistral 7B | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Starling LM 7B Beta | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 35Basic quality (D) · 35 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Granite Code 20B Instruct | Q6_K | 32.6 GB | 43.9 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Codestral Mamba 7B | F16 | 30.6 GB | 40.8 tok/s | Runs great |
| 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Qwen1.5 14B Chat | Q8_0 | 37.7 GB | 45.6 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | Granite 4.0 H Micro | F16 | 22.4 GB | 93.3 tok/s | Runs great |
| 37Basic quality (D) · 37 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. | S5 nano | F16 | 25.5 GB | 63.9 tok/s | Runs great |
| 35Basic quality (D) · 35 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. | OpenChat 3.5 (0106) | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 38Basic quality (D) · 38 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. | DeepSeek LLM 67B Chat | Q4_K_M | 58.7 GB | 18.8 tok/s | Runs well |
| 36Basic quality (D) · 36 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. | Gemma 3n E2B | F16 | 27.0 GB | 149 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Phi-3 Mini 128K Instruct | F16 | 26.8 GB | 77.9 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Voxtral Mini 3B | F16 | 26.3 GB | 63.6 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | DeepSeek Coder 6.7B Instruct | F16 | 33.7 GB | 44.2 tok/s | Runs great |
| 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Yi 1.5 9B Chat | Q8_0 | 25.7 GB | 73.1 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Phi-3 Mini 4K Instruct | F16 | 25.2 GB | 77.9 tok/s | Runs great |
| 36Basic quality (D) · 36 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Qwen2.5-Math 7B Instruct | Q8_0 | 24.3 GB | 84.7 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | Mathstral 7B | F16 | 31.5 GB | 41.0 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | Mistral 7B Instruct v0.2 | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | Nous Hermes 2 Mistral 7B DPO | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | OpenChat 3.5 (1210) | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Falcon-H1 3B Instruct | F16 | 22.5 GB | 94.5 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Qwen2.5 Coder 3B Instruct | F16 | 22.4 GB | 96.3 tok/s | Runs great |
| 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | CodeGemma 7B | Q8_0 | 28.8 GB | 75.6 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | Qwen3-VL 2B Instruct | F16 | 21.2 GB | 140 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | Phi-3 Vision 128K Instruct | F16 | 27.5 GB | 71.7 tok/s | Runs great |
| 35Basic quality (D) · 35 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. | Llama 3.2 3B Instruct | F16 | 23.3 GB | 92.7 tok/s | Runs great |
| 35Basic quality (D) · 35 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. | Raon 1 nano | F16 | 23.2 GB | 84.3 tok/s | Runs great |
| 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | RecurrentGemma 9B IT | Q8_0 | 26.3 GB | 67.1 tok/s | Runs great |
| 32Basic quality (D) · 32 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. | LLM4Decompile 6.7B V2 | F16 | 33.7 GB | 44.2 tok/s | Runs great |
| 32Basic quality (D) · 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | StarCoder2 15B (base) | Q8_0 | 33.5 GB | 40.5 tok/s | Runs great |
| 36Basic quality (D) · 36 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. | MiniCPM5 1B | F16 | 18.3 GB | 276 tok/s | Runs great |
| 34Basic quality (D) · 34 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking. | Aya 23 8B | Q8_0 | 25.5 GB | 80.4 tok/s | Runs great |
| 32Basic quality (D) · 32 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. | OpenHermes 2.5 Mistral 7B | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 32Basic quality (D) · 32 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. | Starling LM 7B Alpha | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | Qwen2.5-VL 3B Instruct | F16 | 23.8 GB | 79.4 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | Qwen2.5 3B Instruct | F16 | 22.4 GB | 96.3 tok/s | Runs great |
| 30Basic quality (D) · 30 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | Code Llama 13B Instruct | Q8_0 | 36.5 GB | 49.6 tok/s | Runs great |
| 34Basic quality (D) · 34 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. | LFM2 2.6B | F16 | 21.2 GB | 116 tok/s | Runs great |
| 34Basic quality (D) · 34 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | Llama 2 70B Chat | Q4_K_M | 59.4 GB | 18.4 tok/s | Runs well |
| 26Basic quality (D) · 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. | DeepSeek MoE 16B Chat | F16 | 49.7 GB | 72.5 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | SQLCoder 7B v2 | F16 | 33.7 GB | 44.2 tok/s | Runs great |
| 32Basic quality (D) · 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | Granite Code 8B Instruct (128K) | Q8_0 | 25.7 GB | 80.2 tok/s | Runs great |
| 32Basic quality (D) · 32 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | InternLM2 Math Plus 7B | Q8_0 | 25.2 GB | 83.4 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | Mistral 7B Instruct v0.1 | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 32Basic quality (D) · 32 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. | Hermes 3 Llama 3.2 3B | F16 | 23.3 GB | 92.7 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | Falcon Mamba 7B Instruct | F16 | 30.5 GB | 40.9 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | Qwen2.5-Omni 3B | F16 | 27.3 GB | 53.7 tok/s | Runs great |
| 32Basic quality (D) · 32 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. | LFM2.5 1.2B Instruct | F16 | 18.4 GB | 254 tok/s | Runs great |
| 28Basic quality (D) · 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | Orca 2 13B | Q8_0 | 33.1 GB | 49.6 tok/s | Runs great |
| 28Basic quality (D) · 28 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. | Neural Chat 7B v3.3 | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 24Basic quality (D) · 24 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. | Qwen1.5 MoE A2.7B Chat | F16 | 46.2 GB | 74.4 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | Falcon 3 3B Instruct | F16 | 23.2 GB | 92.1 tok/s | Runs great |
| 28Basic quality (D) · 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | StableLM 2 12B Chat | Q8_0 | 29.7 GB | 53.2 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | Granite 3.2 2B Instruct | F16 | 21.7 GB | 118 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | Granite 3.3 2B | F16 | 21.7 GB | 118 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | EXAONE 3.5 2.4B | F16 | 21.4 GB | 123 tok/s | Runs great |
| 28Basic quality (D) · 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 56/100 (quality, fit, speed and memory use) sets the ranking. | Gemma 1.1 7B IT | Q8_0 | 28.8 GB | 75.6 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | Qwen3 1.7B | F16 | 21.0 GB | 147 tok/s | Runs great |
| 28Basic quality (D) · 28 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. | Yi 1.5 6B Chat | F16 | 28.3 GB | 49.1 tok/s | Runs great |
| 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | Baichuan 2 13B Chat | Q8_0 | 34.0 GB | 46.5 tok/s | Runs great |
| 30Basic quality (D) · 30 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. | EXAONE 4.0 1.2B | F16 | 19.0 GB | 233 tok/s | Runs great |
| 35Basic quality (D) · 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. | Snowflake Arctic Instruct | Q2_K | 208 GB | 75.3 tok/s | Runs wellover GPU memory limit |
| 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | Llama 2 13B Chat | Q8_0 | 33.1 GB | 49.6 tok/s | Runs great |
| 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | StarCoder 15B | Q8_0 | 32.9 GB | 40.8 tok/s | Runs great |
| 28Basic quality (D) · 28 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | LLaVA 1.6 Mistral 7B | Q8_0 | 25.1 GB | 85.3 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | WizardMath 7B v1.1 | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | Zephyr 7B Beta | F16 | 31.5 GB | 41.1 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | OLMoE 1B-7B Instruct | F16 | 30.3 GB | 114 tok/s | Runs great |
| 28Basic quality (D) · 28 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. | Granite 3.1 2B Instruct | F16 | 21.7 GB | 118 tok/s | Runs great |
| 28Basic quality (D) · 28 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. | Gemma 2 2B | F16 | 21.7 GB | 114 tok/s | Runs great |
| 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | Gemma 7B IT | Q8_0 | 28.8 GB | 75.6 tok/s | Runs great |
| 26Basic quality (D) · 26 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 55/100 (quality, fit, speed and memory use) sets the ranking. | Qwen1.5 7B Chat | Q8_0 | 28.4 GB | 83.6 tok/s | Runs great |
| 28Basic quality (D) · 28 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. | Granite 4.0 H 1B | F16 | 18.9 GB | 204 tok/s | Runs great |
| 24Basic quality (D) · 24 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. | Code Llama 7B Instruct | F16 | 33.7 GB | 44.2 tok/s | Runs great |
| 27Basic quality (D) · 27 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. | WeVisOne VL 2B | F16 | 21.8 GB | 122 tok/s | Runs great |
| 24Basic quality (D) · 24 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. | StarCoder2 7B | F16 | 30.8 GB | 41.5 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | Granite Vision 3.2 2B | F16 | 22.6 GB | 99.9 tok/s | Runs great |
| 23Basic quality (D) · 23 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. | Falcon 180B Chat | Q2_K | 88.1 GB | 11.6 tok/s | Runs well |
| 26Basic quality (D) · 26 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. | Granite 3.0 2B Instruct | F16 | 21.6 GB | 113 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | DeepScaleR 1.5B Preview | F16 | 19.8 GB | 167 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | Qwen2.5 Coder 1.5B Instruct | F16 | 19.3 GB | 193 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | Falcon-H1 1.5B Instruct | F16 | 19.3 GB | 192 tok/s | Runs great |
| 22Basic quality (D) · 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | LLaVA 1.5 13B | Q8_0 | 33.5 GB | 48.4 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | LFM2 1.2B | F16 | 18.4 GB | 254 tok/s | Runs great |
| 22Basic quality (D) · 22 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 53/100 (quality, fit, speed and memory use) sets the ranking. | WizardLM 13B v1.2 | Q8_0 | 33.1 GB | 49.6 tok/s | Runs great |
| 26Basic quality (D) · 26 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. | Qwen3.5 0.8B | F16 | 17.8 GB | 342 tok/s | Runs great |
| 22Basic quality (D) · 22 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. | Orca 2 7B | F16 | 31.6 GB | 44.2 tok/s | Runs great |
| 24Basic quality (D) · 24 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. | SmolVLM2 2.2B Instruct | F16 | 22.1 GB | 132 tok/s | Runs great |
| 24Basic quality (D) · 24 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. | Qwen2-VL 2B Instruct | F16 | 20.6 GB | 135 tok/s | Runs great |
| 24Basic quality (D) · 24 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. | Qwen2.5 1.5B Instruct | F16 | 19.3 GB | 193 tok/s | Runs great |
| 24Basic quality (D) · 24 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. | LFM2-VL 1.6B | F16 | 19.2 GB | 188 tok/s | Runs great |
| 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | Vicuna 13B v1.5 | Q8_0 | 33.1 GB | 49.6 tok/s | Runs great |
| 22Basic quality (D) · 22 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. | Nemotron Mini 4B Instruct | F16 | 24.9 GB | 71.0 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | Llama 2 7B Chat | F16 | 31.6 GB | 44.2 tok/s | Runs great |
| 22Basic quality (D) · 22 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. | SmolVLM 2.2B Instruct | F16 | 22.1 GB | 132 tok/s | Runs great |
| 22Basic quality (D) · 22 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. | Zamba2 2.7B Instruct | F16 | 21.7 GB | 112 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | Yi 6B Chat | F16 | 28.3 GB | 49.1 tok/s | Runs great |
| 22Basic quality (D) · 22 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. | Hunyuan 1.8B Instruct | F16 | 20.1 GB | 166 tok/s | Runs great |
| 22Basic quality (D) · 22 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. | DeepSeek R1 Distill Qwen 1.5B | F16 | 19.8 GB | 167 tok/s | Runs great |
| 22Basic quality (D) · 22 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. | Llama 3.2 1B Instruct | F16 | 18.7 GB | 240 tok/s | Runs great |
| 20Basic quality (D) · 20 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | Baichuan 2 7B Chat | Q8_0 | 26.1 GB | 86.0 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | Granite Code 3B Instruct (128K) | F16 | 25.6 GB | 85.5 tok/s | Runs great |
| 18Basic quality (D) · 18 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. | DeepSeek LLM 7B Chat | F16 | 31.8 GB | 43.1 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | Stable Code Instruct 3B | F16 | 24.2 GB | 106 tok/s | Runs great |
| 21Basic quality (D) · 21 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. | d1 2B | F16 | 19.9 GB | 155 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | Granite 3.1 3B-A800M Instruct | F16 | 23.1 GB | 142 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | SmolLM2 1.7B Instruct | F16 | 21.0 GB | 174 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | Qwen2.5-Math 1.5B Instruct | F16 | 19.2 GB | 193 tok/s | Runs great |
| 20Basic quality (D) · 20 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. | Gemma 3 1B | F16 | 18.0 GB | 298 tok/s | Runs great |
| 16Basic quality (D) · 16 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. | LLaVA 1.5 7B | F16 | 32.2 GB | 42.2 tok/s | Runs great |
| 18Basic quality (D) · 18 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. | Granite 3.0 3B-A800M Instruct | F16 | 23.0 GB | 142 tok/s | Runs great |
| 18Basic quality (D) · 18 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. | StarCoder2 3B | F16 | 22.3 GB | 98.2 tok/s | Runs great |
| 18Basic quality (D) · 18 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. | InternLM2.5 1.8B Chat | F16 | 20.5 GB | 157 tok/s | Runs great |
| 13Basic quality (D) · 13 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. | Falcon 40B Instruct | Q3_K_M | 36.7 GB | 39.2 tok/s | Runs great |
| 18Basic quality (D) · 18 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. | LFM2 700M | F16 | 17.5 GB | 402 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | Vicuna 7B v1.5 | F16 | 31.6 GB | 44.2 tok/s | Runs great |
| 17Basic quality (D) · 17 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. | GLiDE 2B | F16 | 19.8 GB | 158 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | OLMo 7B Instruct | F16 | 30.8 GB | 43.2 tok/s | Runs great |
| 16Basic quality (D) · 16 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. | StableLM Zephyr 3B | F16 | 22.9 GB | 106 tok/s | Runs great |
| 16Basic quality (D) · 16 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. | Phi-2 2.7B | F16 | 22.2 GB | 107 tok/s | Runs great |
| 16Basic quality (D) · 16 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. | Yi Coder 1.5B Chat | F16 | 20.5 GB | 201 tok/s | Runs great |
| 16Basic quality (D) · 16 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. | Falcon 3 1B Instruct | F16 | 19.9 GB | 178 tok/s | Runs great |
| 16Basic quality (D) · 16 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. | Spark X2.6 1.7B | F16 | 19.5 GB | 174 tok/s | Runs great |
| 16Basic quality (D) · 16 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. | Qwen2 1.5B Instruct | F16 | 19.3 GB | 193 tok/s | Runs great |
| 12Basic quality (D) · 12 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. | InternLM 7B | F16 | 31.7 GB | 40.7 tok/s | Runs great |
| 15Basic quality (D) · 15 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. | surya ocr 2.5 | F16 | 17.4 GB | 431 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | CodeGemma 2B | F16 | 21.1 GB | 119 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | SmolLM 1.7B Instruct | F16 | 19.8 GB | 174 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | OLMo 2 1B Instruct | F16 | 19.5 GB | 201 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | Qwen3 0.6B | F16 | 18.4 GB | 397 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | Falcon-H1 0.5B Instruct | F16 | 17.2 GB | 572 tok/s | Runs great |
| 14Basic quality (D) · 14 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. | LFM2.5 350M | F16 | 16.8 GB | 850 tok/s | Runs great |
| 12Basic quality (D) · 12 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. | Gemma 2B IT | F16 | 21.1 GB | 119 tok/s | Runs great |
| 12Basic quality (D) · 12 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. | DeepSeek Coder 1.3B Instruct | F16 | 20.3 GB | 220 tok/s | Runs great |
| 12Basic quality (D) · 12 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. | LLM4Decompile 1.3B V2 | F16 | 20.3 GB | 220 tok/s | Runs great |
| 12Basic quality (D) · 12 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. | LFM2-VL 450M | F16 | 17.0 GB | 661 tok/s | Runs great |
| 12Basic quality (D) · 12 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. | LFM2 350M | F16 | 16.8 GB | 850 tok/s | Runs great |
| 12Basic quality (D) · 12 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. | Granite 4.0 H 350M | F16 | 16.7 GB | 875 tok/s | Runs great |
| 11Basic quality (D) · 11 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. | GLiDE 0.8B | F16 | 17.6 GB | 397 tok/s | Runs great |
| 10Basic quality (D) · 10 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. | StableLM 2 1.6B Chat | F16 | 20.0 GB | 181 tok/s | Runs great |
| 10Basic quality (D) · 10 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. | StableLM 2 Zephyr 1.6B | F16 | 20.0 GB | 181 tok/s | Runs great |
| 10Basic quality (D) · 10 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. | SmolVLM 500M Instruct | F16 | 17.3 GB | 584 tok/s | Runs great |
| 10Basic quality (D) · 10 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. | Qwen2.5 0.5B Instruct | F16 | 17.0 GB | 607 tok/s | Runs great |
| 10Basic quality (D) · 10 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. | LFM2.5 230M | F16 | 16.5 GB | 1,294 tok/s | Runs great |
| 6Basic quality (D) · 6 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. | Falcon 7B Instruct | F16 | 30.4 GB | 41.2 tok/s | Runs great |
| 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking. | Hunyuan 0.5B Instruct | F16 | 17.8 GB | 551 tok/s | Runs great |
| 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking. | Qwen2.5 Coder 0.5B Instruct | F16 | 17.0 GB | 607 tok/s | Runs great |
| 8Basic quality (D) · 8 of 100: how capable the model is at F16, whatever it runs on. Match score 45/100 (quality, fit, speed and memory use) sets the ranking. | SmolLM2 360M Instruct | F16 | 17.0 GB | 827 tok/s | Runs great |
| 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. | TinyLlama 1.1B Chat | F16 | 18.2 GB | 271 tok/s | Runs great |
| 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. | Qwen2 0.5B Instruct | F16 | 17.0 GB | 607 tok/s | Runs great |
| 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. | ERNIE 4.5 0.3B | F16 | 16.8 GB | 827 tok/s | Runs great |
| 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. | SmolVLM 256M Instruct | F16 | 16.7 GB | 1,145 tok/s | Runs great |
| 6Basic quality (D) · 6 of 100: how capable the model is at F16, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. | Gemma 3 270M | F16 | 16.5 GB | 1,102 tok/s | Runs great |
| 5Basic quality (D) · 5 of 100: how capable the model is at F16, whatever it runs on. Match score 44/100 (quality, fit, speed and memory use) sets the ranking. | SmolLM 360M Instruct | F16 | 16.8 GB | 827 tok/s | Runs great |
| 4Basic quality (D) · 4 of 100: how capable the model is at F16, whatever it runs on. Match score 43/100 (quality, fit, speed and memory use) sets the ranking. | SmolLM2 135M Instruct | F16 | 16.4 GB | 2,289 tok/s | Runs great |
| 3Basic quality (D) · 3 of 100: how capable the model is at F16, whatever it runs on. Match score 43/100 (quality, fit, speed and memory use) sets the ranking. | Falcon-H1 Tiny 90M Instruct | F16 | 16.2 GB | 3,307 tok/s | Runs great |
| 2Basic quality (D) · 2 of 100: how capable the model is at F16, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking. | SmolLM 135M Instruct | F16 | 16.3 GB | 2,289 tok/s | Runs great |
| 69Good quality (B) · 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. | Kimi K3.1 Pro (2.8T) | Q2_K | 1,128 GB | — | Won't run |
| 68Good quality (B) · 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. | DeepSeek V4.1 Flash 1017 (552B-A16B) | Q2_K | 321 GB | — | Won't run |
| 68Good quality (B) · 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. | GLM-5.3 (753B-A40B) | Q2_K | 318 GB | — | Won't run |
| 68Good quality (B) · 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. | Qwen3.8 Max (2.4T-A95B) | Q2_K | 995 GB | — | Won't run |
| 68Good quality (B) · 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. | Kimi K3 (2.8T) | Q2_K | 1,128 GB | — | Won't run |
| 68Good quality (B) · 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. | LongCat 2.5 (1.6T-A48B) | Q2_K | 728 GB | — | Won't run |
| 68Good quality (B) · 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. | MiMo V2.6 Pro (1T-A42B) | Q2_K | 426 GB | — | Won't run |
| 68Good quality (B) · 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. | Ling 3.1 Flash 1019 (560B-A25B) | Q2_K | 245 GB | — | Won't run |
| 67Good quality (B) · 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. | Step 5 Preview (600B-A27B) | Q2_K | 258 GB | — | Won't run |
| 67Good quality (B) · 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. | DeepSeek V4.1 Flash (552B-A16B) | Q2_K | 321 GB | — | Won't run |
| 67Good quality (B) · 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. | Mistral Large 4 (1T-A52B) | Q2_K | 438 GB | — | Won't run |
| 67Good quality (B) · 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. | GLM-5.2 (753B-A40B) | Q2_K | 318 GB | — | Won't run |
| 67Good quality (B) · 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. | Hunyuan Hy4 Preview (780B) | Q2_K | 329 GB | — | Won't run |
| 67Good quality (B) · 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. | Muse Spark Open (816B-A42B) | Q2_K | 345 GB | — | Won't run |
| 67Good quality (B) · 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. | DeepSeek V4 Pro (1.6T-A49B) | Q2_K | 677 GB | — | Won't run |
| 66Good quality (B) · 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. | Kimi K2.6 (1T) | Q2_K | 427 GB | — | Won't run |
| 66Good quality (B) · 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. | Ling 3.1 Flash (560B-A25B) | Q2_K | 245 GB | — | Won't run |
| 66Good quality (B) · 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. | MiMo V2.5 Pro (1T-A42B) | Q2_K | 426 GB | — | Won't run |
| 66Good quality (B) · 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. | GLM-5.1 (744B-A40B) | Q2_K | 318 GB | — | Won't run |
| 66Good quality (B) · 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. | DeepSeek V4 Pro (April 2026) | Q2_K | 656 GB | — | Won't run |
| 66Good quality (B) · 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. | GLM-5 (744B-A40B) | Q2_K | 318 GB | — | Won't run |
| 66Good quality (B) · 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. | Kimi K2.5 (1T) | Q2_K | 427 GB | — | Won't run |
| 66Good quality (B) · 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. | Kimi K2.7 Code (1T) | Q2_K | 427 GB | — | Won't run |
| 65Good quality (B) · 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. | DeepSeek V3.2 (671B) | Q2_K | 291 GB | — | Won't run |
| 65Good quality (B) · 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. | Inkling (975B-A41B) | Q2_K | 397 GB | — | Won't run |
| 65Good quality (B) · 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. | Nemotron 3 Ultra 550B-A55B | Q2_K | 240 GB | — | Won't run |
| 65Good quality (B) · 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. | DeepSeek R1 0528 (671B) | Q2_K | 290 GB | — | Won't run |
| 65Good quality (B) · 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. | Kimi K2 Thinking (1T) | Q2_K | 427 GB | — | Won't run |
| 64Good quality (B) · 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. | DeepSeek V3.2 Exp | Q2_K | 291 GB | — | Won't run |
| 64Good quality (B) · 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. | Ling 2.6 1T | Q2_K | 426 GB | — | Won't run |
| 64Good quality (B) · 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. | DeepSeek R1 (671B) | Q2_K | 285 GB | — | Won't run |
| 64Good quality (B) · 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. | DeepSeek V3.1 Terminus | Q2_K | 290 GB | — | Won't run |
| 63Good quality (B) · 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. | DeepSeek V3.1 (671B) | Q2_K | 290 GB | — | Won't run |
| 63Good quality (B) · 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. | Kimi K2 Instruct 0905 (1T) | Q2_K | 427 GB | — | Won't run |
| 63Good quality (B) · 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. | Ling 1T | Q2_K | 419 GB | — | Won't run |
| 63Good quality (B) · 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. | Mistral Large 3 (675B) | Q2_K | 287 GB | — | Won't run |
| 62Good quality (B) · 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. | Kimi K2 Instruct (July 2025) | Q2_K | 427 GB | — | Won't run |
| 61Good quality (B) · 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. | LongCat Flash Chat (560B) | Q2_K | 241 GB | — | Won't run |
| 58Good quality (B) · 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. | DeepSeek V3 0324 (671B) | Q2_K | 290 GB | — | Won't run |
| 52Fair quality (C) · 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. | DeepSeek V3 (671B) | Q2_K | 285 GB | — | Won't run |
Image models on the Apple M5 Ultra (256 GB)
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 93Excellent quality (A+) · 93 of 100: how capable the model is at FP16, whatever it runs on. Match score 89/100 (quality, fit, speed and memory use) sets the ranking. | MiniMax H3 Image | FP16 | 138 GB | 34 s | Runs well |
| 92Excellent quality (A+) · 92 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. | FLUX.2 [dev] | FP16 | 116 GB | 2.1 min | Slow |
| 92Excellent quality (A+) · 92 of 100: how capable the model is at FP16, whatever it runs on. Match score 83/100 (quality, fit, speed and memory use) sets the ranking. | Krea 2 Turbo | FP16 | 37.4 GB | 15 s | Runs well |
| 84Great quality (A) · 84 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. | FLUX.2 [klein] 9B | FP16 | 37.1 GB | 6.2 s | Runs great |
| 85Excellent quality (A+) · 85 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. | ERNIE-Image Turbo | FP16 | 26.4 GB | 11 s | Runs great |
| 84Great quality (A) · 84 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. | Z-Image Turbo | FP16 | 22.7 GB | 8.6 s | Runs great |
| 91Excellent quality (A+) · 91 of 100: how capable the model is at FP16, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking. | FLUX 3 Image Klein | FP16 | 22.4 GB | 20 s | Runs well |
| 88Excellent quality (A+) · 88 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. | HunyuanImage 3.0 (80B-A13B) | FP8 | 86.9 GB | 3.1 min | Very slow |
| 92Excellent quality (A+) · 92 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. | Ideogram 4 | FP16 | 57.3 GB | 86 s | Slow |
| 89Excellent quality (A+) · 89 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. | Qwen-Image 2.1 Turbo | FP16 | 46.7 GB | 40 s | Runs well |
| 90Excellent quality (A+) · 90 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. | Cosmos3-Super Text2Image (64B) | FP8 | 69.1 GB | 2.1 min | Slow |
| 86Excellent quality (A+) · 86 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. | Ming-Image 0.1 Design | FP16 | 63.9 GB | 56 s | Slow |
| 78Great quality (A) · 78 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. | FLUX.2 [klein] 4B | FP16 | 18.1 GB | 3.4 s | Runs great |
| 93Excellent quality (A+) · 93 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. | Qwen-Image 2.1 | FP16 | 57.0 GB | 3.4 min | Very slow |
| 90Excellent quality (A+) · 90 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. | Qwen-Image 2512 | FP16 | 63.8 GB | 5.1 min | Very slow |
| 70Great quality (A) · 70 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. | FLUX.1 [schnell] | FP16 | 36.4 GB | 6.8 s | Runs great |
| 93Excellent quality (A+) · 93 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. | Fibo 2.0 Dev | FP16 | 27.7 GB | 77 s | Slow |
| 76Great quality (A) · 76 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. | HiDream-I1 Fast | FP16 | 64.4 GB | 34 s | Runs well |
| 88Excellent quality (A+) · 88 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. | Qwen-Image | FP16 | 63.8 GB | 5.1 min | Very slow |
| 80Great quality (A) · 80 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. | HiDream-I1 Dev | FP16 | 64.4 GB | 58 s | Slow |
| 87Excellent quality (A+) · 87 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. | FireRed-Image-Edit 1.1 | FP16 | 61.2 GB | 4.6 min | Very slow |
| 84Great quality (A) · 84 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. | HiDream-I1 Full | FP16 | 64.4 GB | 2.6 min | Very slow |
| 88Excellent quality (A+) · 88 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. | ERNIE-Image | FP16 | 34.0 GB | 75 s | Slow |
| 86Excellent quality (A+) · 86 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. | Qwen-Image-Edit 2511 | FP16 | 61.2 GB | 4.6 min | Very slow |
| 82Great quality (A) · 82 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. | MOSS-Image SFT | FP16 | 29.3 GB | 48 s | Slow |
| 82Great quality (A) · 82 of 100: how capable the model is at FP16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | FLUX.1 Kontext [dev] | FP16 | 37.0 GB | 75 s | Slow |
| 62Good quality (B) · 62 of 100: how capable the model is at FP16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | Stable Diffusion 3.5 Large Turbo | FP16 | 30.0 GB | 5.5 s | Runs great |
| 80Great quality (A) · 80 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. | FLUX.1 Krea [dev] | FP8 | 19.6 GB | 37 s | Runs well |
| 82Great quality (A) · 82 of 100: how capable the model is at FP16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | Z-Image | FP16 | 22.7 GB | 58 s | Slow |
| 78Great quality (A) · 78 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. | FLUX.1 [dev] | FP8 | 19.6 GB | 37 s | Runs well |
| 78Great quality (A) · 78 of 100: how capable the model is at FP16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | Chroma1-HD | FP16 | 30.0 GB | 68 s | Slow |
| 74Great quality (A) · 74 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. | FLUX.1 Fill [dev] | FP8 | 19.6 GB | 38 s | Runs well |
| 58Good quality (B) · 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | DreamShaper XL v2 Turbo | FP16 | 8.9 GB | 4.4 s | Runs great |
| 70Great quality (A) · 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | CogView4 6B | FP16 | 33.9 GB | 62 s | Slow |
| 58Good quality (B) · 58 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. | Kolors | FP8 | 11.0 GB | 12 s | Runs great |
| 66Good quality (B) · 66 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | Stable Diffusion 3.5 Large | FP16 | 30.0 GB | 38 s | Runs well |
| 72Great quality (A) · 72 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. | Iris-3B | FP16 | 17.3 GB | 55 s | Slow |
| 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | Juggernaut XL v9 | FP8 | 5.6 GB | 14 s | Runs well |
| 60Good quality (B) · 60 of 100: how capable the model is at FP8, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | RealVisXL V5.0 | FP8 | 5.6 GB | 14 s | Runs well |
| 54Fair quality (C) · 54 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. | SDXL-Lightning (4-step) | FP16 | 8.9 GB | 2.4 s | Runs great |
| 58Good quality (B) · 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | Animagine XL 4.0 | FP8 | 5.6 GB | 13 s | Runs great |
| 58Good quality (B) · 58 of 100: how capable the model is at FP8, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | Illustrious XL v0.1 | FP8 | 5.6 GB | 13 s | Runs great |
| 56Good quality (B) · 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 65/100 (quality, fit, speed and memory use) sets the ranking. | Pony Diffusion V6 XL | FP8 | 5.6 GB | 12 s | Runs great |
| 52Fair quality (C) · 52 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. | Stable Diffusion 3 Medium | FP16 | 17.4 GB | 11 s | Runs great |
| 62Good quality (B) · 62 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. | Lumina-Image 2.0 | FP16 | 12.7 GB | 24 s | Runs well |
| 50Fair quality (C) · 50 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. | PixArt-Σ XL 1024 | FP16 | 12.8 GB | 5.1 s | Runs great |
| 54Fair quality (C) · 54 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. | Animagine XL 3.1 | FP8 | 5.6 GB | 13 s | Runs great |
| 66Good quality (B) · 66 of 100: how capable the model is at FP16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | OmniGen2 | FP16 | 18.0 GB | 55 s | Slow |
| 56Good quality (B) · 56 of 100: how capable the model is at FP8, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Stable Diffusion 3.5 Medium | FP8 | 10.4 GB | 18 s | Runs well |
| 53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Stable Diffusion XL 1.0 | FP8 | 5.6 GB | 14 s | Runs well |
| 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Playground v2.5 | FP16 | 8.9 GB | 26 s | Runs well |
| 50Fair quality (C) · 50 of 100: how capable the model is at FP8, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | Stable Cascade | FP8 | 8.1 GB | 15 s | Runs well |
| 58Good quality (B) · 58 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. | AuraFlow v0.3 | FP16 | 18.5 GB | 48 s | Slow |
| 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Hunyuan-DiT v1.2 | FP16 | 9.3 GB | 21 s | Runs well |
| 42Fair quality (C) · 42 of 100: how capable the model is at FP16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Realistic Vision V5.1 | FP16 | 2.8 GB | 4.1 s | Runs great |
| 40Fair quality (C) · 40 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. | SDXL Turbo | FP16 | 7.6 GB | 0.5 s | Runs great |
| 40Fair quality (C) · 40 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. | DreamShaper 8 | FP16 | 2.8 GB | 4.1 s | Runs great |
| 48Fair quality (C) · 48 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. | Kandinsky 3 | FP16 | 25.9 GB | 31 s | Runs well |
| 35Basic quality (D) · 35 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. | Stable Diffusion 1.5 | FP16 | 2.8 GB | 4.1 s | Runs great |
| 30Basic quality (D) · 30 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. | LCM DreamShaper v7 | FP16 | 2.8 GB | 0.9 s | Runs great |
| 28Basic quality (D) · 28 of 100: how capable the model is at FP16, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking. | SD Turbo | FP16 | 3.3 GB | 0.3 s | Runs great |
Video models on the Apple M5 Ultra (256 GB)
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 88Excellent quality (A+) · 88 of 100: how capable the model is at FP16, whatever it runs on. Match score 85/100 (quality, fit, speed and memory use) sets the ranking. | LTX-2.5 (22B, video + audio) | FP16 | 82.5 GB | 5.2 min | Runs well |
| 90Excellent quality (A+) · 90 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. | MiniMax H3 (33B, video + audio) | FP16 | 113 GB | 59.3 min | Very slow |
| 88Excellent quality (A+) · 88 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. | Cosmos3-Super Image2Video (64B) | FP16 | 139 GB | 2.1 h | Very slow |
| 86Excellent quality (A+) · 86 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. | LTX-2.3 | FP16 | 75.4 GB | 22.8 min | Slow |
| 88Excellent quality (A+) · 88 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. | Kandinsky 6.0 Pro (29B, video + audio) | FP16 | 85.4 GB | 2.8 h | Very slow |
| 82Great quality (A) · 82 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. | LTX-2 (19B, video + audio) | FP16 | 75.4 GB | 22.8 min | Slow |
| 66Good quality (B) · 66 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. | LTX-Video 13B 0.9.8 (distilled) | FP8 | 30.1 GB | 3.1 min | Runs great |
| 86Excellent quality (A+) · 86 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. | Wan 2.2 I2V A14B | FP16 | 58.0 GB | 1.5 h | Very slow |
| 86Excellent quality (A+) · 86 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. | Wan 2.2 T2V A14B | FP16 | 58.0 GB | 1.5 h | Very slow |
| 84Great quality (A) · 84 of 100: how capable the model is at FP16, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. | HunyuanVideo 1.5 (8.3B) | FP16 | 55.2 GB | 1.7 h | Very slow |
| 76Great quality (A) · 76 of 100: how capable the model is at FP16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | Step-Video T2V (30B) | FP16 | 78.2 GB | 2.9 h | Very slow |
| 80Great quality (A) · 80 of 100: how capable the model is at FP16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | SCAIL-2 14B (character animation) | FP16 | 52.5 GB | 68.4 min | Very slow |
| 78Great quality (A) · 78 of 100: how capable the model is at FP16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.2 Animate 14B | FP16 | 65.2 GB | 1.8 h | Very slow |
| 78Great quality (A) · 78 of 100: how capable the model is at FP16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.1 VACE 14B | FP16 | 64.6 GB | 2.1 h | Very slow |
| 76Great quality (A) · 76 of 100: how capable the model is at FP16, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking. | Wan-Dancer 14B (music to dance) | FP16 | 77.0 GB | 4.2 h | Very slow |
| 83Great quality (A) · 83 of 100: how capable the model is at FP8, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | Bernini-R A14B (video generation + editing) | FP8 | 26.0 GB | 37.0 min | Slow |
| 74Great quality (A) · 74 of 100: how capable the model is at FP16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | FramePack I2V (HunyuanVideo 13B) | FP16 | 49.6 GB | 22.8 min | Slow |
| 78Great quality (A) · 78 of 100: how capable the model is at FP16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.1 I2V 14B 720P | FP16 | 59.3 GB | 1.9 h | Very slow |
| 78Great quality (A) · 78 of 100: how capable the model is at FP16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | SkyReels V2 T2V 14B 720P | FP16 | 60.8 GB | 2.3 h | Very slow |
| 78Great quality (A) · 78 of 100: how capable the model is at FP16, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.1 T2V 14B | FP16 | 58.0 GB | 1.9 h | Very slow |
| 76Great quality (A) · 76 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. | HY-OmniWeaving (8.3B, unified video) | FP16 | 41.9 GB | 30.2 min | Slow |
| 76Great quality (A) · 76 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. | Wan 2.2 S2V 14B (speech to video) | FP16 | 50.4 GB | 57.0 min | Very slow |
| 74Great quality (A) · 74 of 100: how capable the model is at FP16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.1 I2V 14B 480P | FP16 | 46.5 GB | 51.3 min | Very slow |
| 72Great quality (A) · 72 of 100: how capable the model is at FP16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | HunyuanVideo 13B | FP16 | 64.8 GB | 2.3 h | Very slow |
| 72Great quality (A) · 72 of 100: how capable the model is at FP16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | HunyuanVideo I2V 13B | FP16 | 64.8 GB | 2.3 h | Very slow |
| 74Great quality (A) · 74 of 100: how capable the model is at FP16, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | SkyReels V2 T2V 14B 540P | FP16 | 49.0 GB | 79.8 min | Very slow |
| 70Great quality (A) · 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.2 TI2V 5B | FP16 | 43.1 GB | 29.7 min | Slow |
| 70Great quality (A) · 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. | Cosmos-Predict2 14B Video2World | FP16 | 57.6 GB | 1.5 h | Very slow |
| 68Good quality (B) · 68 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. | Kandinsky 6.0 Video Lite (3B, video + audio) | FP16 | 31.5 GB | 17.7 min | Slow |
| 50Fair quality (C) · 50 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. | LTX-Video 2B | FP16 | 17.5 GB | 86 s | Runs great |
| 70Great quality (A) · 70 of 100: how capable the model is at FP16, whatever it runs on. Match score 63/100 (quality, fit, speed and memory use) sets the ranking. | Open-Sora 2.0 (11B) | FP16 | 43.8 GB | 2.9 h | Very slow |
| 60Good quality (B) · 60 of 100: how capable the model is at FP16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Cosmos 1.0 Diffusion 7B Text2World | FP16 | 48.2 GB | 31.4 min | Slow |
| 58Good quality (B) · 58 of 100: how capable the model is at FP16, whatever it runs on. Match score 62/100 (quality, fit, speed and memory use) sets the ranking. | Cosmos-Predict2 2B Video2World | FP16 | 33.6 GB | 14.3 min | Slow |
| 60Good quality (B) · 60 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. | CogVideoX 1.5 5B | FP16 | 50.3 GB | 51.3 min | Very slow |
| 62Good quality (B) · 62 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. | Mochi 1 Preview (10B) | FP16 | 41.4 GB | 62.7 min | Very slow |
| 44Fair quality (C) · 44 of 100: how capable the model is at FP8, whatever it runs on. Match score 60/100 (quality, fit, speed and memory use) sets the ranking. | CogVideoX 2B | FP8 | 11.0 GB | 3.1 min | Runs great |
| 55Good quality (B) · 55 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. | CogVideoX 5B I2V | FP16 | 25.3 GB | 10.3 min | Runs well |
| 55Good quality (B) · 55 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. | CogVideoX 5B | FP16 | 24.1 GB | 10.3 min | Runs well |
| 55Good quality (B) · 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.1 T2V 1.3B | FP16 | 19.8 GB | 13.7 min | Slow |
| 55Good quality (B) · 55 of 100: how capable the model is at FP16, whatever it runs on. Match score 59/100 (quality, fit, speed and memory use) sets the ranking. | Wan 2.1 VACE 1.3B | FP16 | 21.5 GB | 14.8 min | Slow |
| 52Fair quality (C) · 52 of 100: how capable the model is at FP16, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking. | SkyReels V2 Diffusion Forcing 1.3B | FP16 | 23.8 GB | 17.1 min | Slow |
| 35Basic quality (D) · 35 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. | AnimateDiff v3 (SD 1.5) | FP16 | 4.7 GB | 68 s | Runs great |
| 52Fair quality (C) · 52 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. | Allegro (2.8B) | FP16 | 38.2 GB | 68.4 min | Very slow |
Decision models on the Apple M5 Ultra (256 GB)
| Quality | Model | Quant | Memory | Speed | Runs |
|---|---|---|---|---|---|
| 90Excellent quality (A+) · 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. | Qwen3 Reranker 4B | Q8_0 | 20.5 GB | 98 ms | Runs great |
| 88Excellent quality (A+) · 88 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 94/100 (quality, fit, speed and memory use) sets the ranking. | zerank-2 | Q8_0 | 20.5 GB | 98 ms | Runs great |
| 86Excellent quality (A+) · 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. | Jina Reranker v3.5 | F16 | 17.4 GB | 19 ms | Runs great |
| 86Excellent quality (A+) · 86 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. | Shieldstral 1.0 3B | Q8_0 | 20.3 GB | 94 ms | Runs great |
| 84Great quality (A) · 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. | Intern-Decision 4B | F16 | 25.1 GB | 69 ms | Runs great |
| 84Great quality (A) · 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. | Jina Reranker v3 | F16 | 17.4 GB | 19 ms | Runs great |
| 82Great quality (A) · 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. | d1 3B | F16 | 22.2 GB | 49 ms | Runs great |
| 82Great quality (A) · 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. | mxbai-rerank-large v2 | F16 | 19.1 GB | 43 ms | Runs great |
| 90Excellent quality (A+) · 90 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. | Qwen3 Reranker 8B | Q8_0 | 24.9 GB | 196 ms | Runs well |
| 80Great quality (A) · 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. | FRIDA-Decisions | F16 | 2.4 GB | 15 ms | Runs great |
| 80Great quality (A) · 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. | GLiNER2.5 Decide | F16 | 1.7 GB | 11 ms | Runs great |
| 80Great quality (A) · 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. | LemonDecider 4B | F16 | 10.8 GB | 66 ms | Runs great |
| 80Great quality (A) · 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. | Qwen3 Reranker 0.6B | F16 | 17.4 GB | 19 ms | Runs great |
| 80Great quality (A) · 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. | zerank-1 small | F16 | 19.6 GB | 47 ms | Runs great |
| 78Great quality (A) · 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. | BGE Reranker v2 Gemma | F16 | 21.0 GB | 67 ms | Runs great |
| 76Great quality (A) · 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. | Llama Prompt Guard 2 86M | F16 | 1.2 GB | 12 ms | Runs great |
| 76Great quality (A) · 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. | mxbai-rerank-base v2 | F16 | 17.0 GB | 16 ms | Runs great |
| 90Excellent quality (A+) · 90 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. | Skywork Reward V2 Llama 3.1 8B | Q8_0 | 10.4 GB | 361 ms | Runs well |
| 74Great quality (A) · 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. | GLiClass ModernBERT Large v3 | F16 | 1.5 GB | 9.5 ms | Runs great |
| 78Great quality (A) · 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. | OpenAI Privacy Filter | F16 | 3.8 GB | 4.7 ms | Runs great |
| 83Great quality (A) · 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. | Qwen3Guard Gen 4B | Q6_K | 19.9 GB | 131 ms | Runs great |
| 72Great quality (A) · 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. | BGE Reranker v2 M3 | F16 | 1.9 GB | 20 ms | Runs great |
| 72Great quality (A) · 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. | DeBERTa v3 Large Zero-shot v2.0 | F16 | 1.6 GB | 10 ms | Runs great |
| 72Great quality (A) · 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. | DuoGuard 0.5B | F16 | 1.7 GB | 17 ms | Runs great |
| 72Great quality (A) · 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. | Skywork Reward V2 Qwen3 0.6B | F16 | 2.0 GB | 37 ms | Runs great |
| 86Excellent quality (A+) · 86 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. | GLiDE 27B | Q8_0 | 34.8 GB | 315 ms | Runs well |
| 70Great quality (A) · 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. | GTE Reranker ModernBERT Base | F16 | 0.9 GB | 8.1 ms | Runs great |
| 70Great quality (A) · 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. | Jina Reranker v2 Base Multilingual | F16 | 1.2 GB | 12 ms | Runs great |
| 70Great quality (A) · 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. | ModernBERT Large Zero-shot v2.0 | F16 | 1.5 GB | 9.5 ms | Runs great |
| 84Great quality (A) · 84 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. | Skywork Reward V2 Qwen3 8B | Q8_0 | 10.5 GB | 364 ms | Runs well |
| 68Good quality (B) · 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. | GLiGuard 300M | F16 | 1.1 GB | 9.7 ms | Runs great |
| 68Good quality (B) · 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. | Granite Embedding Reranker English R2 | F16 | 0.9 GB | 8.1 ms | Runs great |
| 68Good quality (B) · 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. | Laya | F16 | 1.5 GB | 9.7 ms | Runs great |
| 68Good quality (B) · 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. | Llama Prompt Guard 2 22M | F16 | 0.8 GB | 5.9 ms | Runs great |
| 74Great quality (A) · 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. | Qwen3Guard Gen 0.6B | F16 | 17.7 GB | 36 ms | Runs great |
| 86Excellent quality (A+) · 86 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. | Qwen3Guard Gen 8B | Q8_0 | 24.9 GB | 259 ms | Runs well |
| 66Good quality (B) · 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. | GLiClass Base v3 | F16 | 1.0 GB | 6.6 ms | Runs great |
| 66Good quality (B) · 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. | ProtectAI DeBERTa v3 Prompt Injection v2 | F16 | 1.0 GB | 8.9 ms | Runs great |
| 72Great quality (A) · 72 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. | NuExtract 2.0 4B | Q4_K_M | 18.3 GB | 116 ms | Runs great |
| 83Great quality (A) · 83 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. | Llama 3.1 Nemotron Safety Guard 8B v3 | Q6_K | 22.8 GB | 235 ms | Runs well |
| 69Good quality (B) · 69 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. | NuExtract 2.0 2B | Q6_K | 17.8 GB | 88 ms | Runs great |
| 83Great quality (A) · 83 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. | Granite Guardian 4.1 8B | Q6_K | 23.2 GB | 245 ms | Runs well |
| 64Good quality (B) · 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. | DeBERTa v3 Base Zero-shot v2.0 | F16 | 1.0 GB | 6.5 ms | Runs great |
| 68Good quality (B) · 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. | GLiNER2 Base v1 | F16 | 1.1 GB | 6.9 ms | Runs great |
| 79Great quality (A) · 79 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. | WildGuard 7B | Q6_K | 22.2 GB | 212 ms | Runs well |
| 62Good quality (B) · 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. | HHEM 2.1 Open | F16 | 0.8 GB | 10 ms | Runs great |
| 66Good quality (B) · 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. | LettuceDetect Large | F16 | 1.5 GB | 26 ms | Runs great |
| 79Great quality (A) · 79 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. | Granite Guardian 3.3 8B | Q6_K | 23.0 GB | 239 ms | Runs well |
| 60Good quality (B) · 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. | Laya Multilingual | F16 | 1.3 GB | 8.4 ms | Runs great |
| 58Good quality (B) · 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. | ShieldGemma 2B | F16 | 21.4 GB | 70 ms | Runs great |
| 62Good quality (B) · 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. | GLiNER Multi v2.1 | F16 | 1.3 GB | 8.0 ms | Runs great |
| 56Good quality (B) · 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. | BART Large MNLI | F16 | 1.5 GB | 9.6 ms | Runs great |
| 64Good quality (B) · 64 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. | ShieldGemma 9B | Q8_0 | 26.5 GB | 221 ms | Runs well |
| 52Fair quality (C) · 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. | ms-marco MiniLM L6 v2 | F16 | 0.7 GB | 4.6 ms | Runs great |
| 56Good quality (B) · 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. | Llama Guard 3 1B | F16 | 19.0 GB | 67 ms | Runs great |
| 67Good quality (B) · 67 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. | Llama Guard 3 8B | Q6_K | 22.8 GB | 235 ms | Runs well |
| 66Good quality (B) · 66 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. | ShieldGemma 27B | Q8_0 | 45.6 GB | 642 ms | Slow |
| 80Great quality (A) · 80 of 100: how capable the model is at MXFP4, whatever it runs on. Match score 75/100 (quality, fit, speed and memory use) sets the ranking. | gpt-oss-safeguard 120B | MXFP4 | 82.6 GB | 1.3 s | Slow |
| 77Great quality (A) · 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. | Nemotron Content Safety Reasoning 4B | Q6_K | 19.8 GB | 880 ms | Slow |
| 65Good quality (B) · 65 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. | Llama Guard 4 12B | Q6_K | 26.2 GB | 349 ms | Runs well |
| 86Excellent quality (A+) · 86 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. | CompassJudger-2 32B | Q8_0 | 51.2 GB | 9.0 s | Very slow |
| 76Great quality (A) · 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. | gpt-oss-safeguard 20B | MXFP4 | 28.3 GB | 1.0 s | Slow |
| 66Good quality (B) · 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. | DynaGuard 1.7B | Q4_K_M | 17.4 GB | 305 ms | Runs well |
| 75Great quality (A) · 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. | DynaGuard 8B | Q4_K_M | 21.3 GB | 1.2 s | Slow |
| 76Great quality (A) · 76 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. | Selene 1 Mini 8B | Q4_K_M | 21.1 GB | 1.3 s | Slow |
| 74Great quality (A) · 74 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. | CompassJudger-2 7B | Q4_K_M | 20.7 GB | 1.2 s | Slow |
| 74Great quality (A) · 74 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. | Prometheus 2 8x7B | Q8_0 | 65.7 GB | 4.1 s | Very slow |
| 65Good quality (B) · 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. | GLIDER | Q6_K | 19.9 GB | 822 ms | Slow |
| 63Good quality (B) · 63 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | Flow Judge v0.1 | Q6_K | 19.9 GB | 822 ms | Slow |
| 64Good quality (B) · 64 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. | Prometheus 2 7B | Q4_K_M | 20.6 GB | 1.2 s | Slow |
Upgrade picks
No consumer GPU in our catalog runs more LLMs well than your current setup.
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 “Measured” come from a linked benchmark.