Hardware · Rockchip

What AI models can the Rockchip RK3566 / RK3568 (8 GB) run?

RAM: 8.0 GB
Usable for models: 6.4 GB
Memory bandwidth: 13 GB/s
Runs models on: CPU cores
Released: 2021

288 / 622 LLMs run Best: Nemotron Labs Dory 1.0 4B 44 / 61 Image models run Best: FLUX 3 Image Klein 6 / 43 Video models run Best: AnimateDiff v3 (SD 1.5) 59 / 70 Decision models run Best: Qwen3 Reranker 4B

Models run on the CPU cores here, so the memory’s speed and how fast those cores can read it set the pace: small models (about 1-4B) are usable, bigger ones are slow. We plan around 6.4 GB of the 8.0 GB (what Linux leaves free for a model). 0.8 TOPS NPU (about 1 TOPS on the RK3568) for small vision models through Rockchip’s RKNN toolkit.

More about this hardware

Maker	Rockchip
Type	Single-board computer or edge AI board
Share of memory for models	80% (what Linux typically leaves free on a board this size)
Chip	Rockchip RK3566
AI accelerator	0.8 TOPS NPU (about 1 TOPS on the RK3568) for small vision models through Rockchip’s RKNN toolkit.
Boards with this chip	Radxa ROCK 3A / 3C / ZERO 3 · Orange Pi 3B · Pine64 Quartz64 · FriendlyElec NanoPi R5S
Memory type	LPDDR4/LPDDR4X or DDR4 up to 3200 MT/s, 32-bit
Bandwidth figure	12.8 GB/s, the chip’s rated memory speed
Note	4× Cortex-A55 at 1.8 GHz (RK3566) or 2.0 GHz (RK3568), Mali-G52 2EE GPU, up to 8 GB. Entry level: small models only, at a few tokens per second. Four A55 cores can’t read memory at anything near its rated speed, so we assume 20% (estimate).
Specifications	www.cnx-software.com

Apps that run models well on the Rockchip RK3566 / RK3568 (8 GB)

Linux on ARM boards

For a typical model that fits in memory. Each model page picks the best app for that model on this hardware. All apps

Best LLMs for the Rockchip RK3566 / RK3568 (8 GB)

Quality	Model	Quant	Memory	Speed	Runs
67Good quality (B) · 67 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.	Nemotron Labs Dory 1.0 4BNVIDIA	`Q6_K`	4.8 GB	0.8 tok/s	Very slow
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.	Nanbeige4.2 3BBOSS Zhipin (Nanbeige)	`Q4_K_M`	5.1 GB	1.0 tok/s	Very slow
65Good quality (B) · 65 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.	Spark-X2.6 4BSparkLLM	`Q6_K`	4.8 GB	0.7 tok/s	Very slow
66Good quality (B) · 66 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.	Nemotron 3.5 Bolt E4BNVIDIA	`Q4_K_M`	6.1 GB	1.0 tok/s	Very slowtight fit
64Good quality (B) · 64 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.	Gemma 4 E4BGoogle	`Q4_K_M`	6.1 GB	1.0 tok/s	Very slowtight fit
65Good quality (B) · 65 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking.	MiMo V2.6 Distill Qwen 9BXiaomi	`Q3_K_M`	6.0 GB	0.6 tok/s	Very slowtight fit
65Good quality (B) · 65 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 70/100 (quality, fit, speed and memory use) sets the ranking.	SenseNova Loop 9B PreviewSenseTime	`Q3_K_M`	6.3 GB	0.6 tok/s	Very slowtight fit
64Good quality (B) · 64 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	Ornith 1.5 9BDeepReinforce	`Q3_K_M`	6.1 GB	0.6 tok/s	Very slowtight fit
63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	Qwen3.5 9BAlibaba	`Q3_K_M`	6.1 GB	0.6 tok/s	Very slowtight fit
61Good quality (B) · 61 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	Qwen3.5 4BAlibaba	`Q6_K`	5.2 GB	0.6 tok/s	Very slow
60Good quality (B) · 60 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	TMax 4BAi2	`Q6_K`	4.8 GB	0.7 tok/s	Very slow
63Good quality (B) · 63 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	Ornith 1.0 9BDeepReinforce	`Q3_K_M`	6.1 GB	0.6 tok/s	Very slowtight fit

Best Image models for the Rockchip RK3566 / RK3568 (8 GB)

Quality	Model	Quant	Memory	Speed	Runs
84Great quality (A) · 84 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 72/100 (quality, fit, speed and memory use) sets the ranking.	FLUX 3 Image KleinBlack Forest Labs	`Q4 GGUF`	5.7 GB	43.3 min	Very slowtight fit · encoder unloaded after use
76Great quality (A) · 76 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.	FLUX.2 [klein] 4BBlack Forest Labs	`FP8`	6.2 GB	5.7 min	Very slowtight fit · encoder unloaded after use
76Great quality (A) · 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	MOSS-Image SFTOpenMOSS	`Q4 GGUF`	5.4 GB	1.8 h	Very slowencoder unloaded after use
78Great quality (A) · 78 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 69/100 (quality, fit, speed and memory use) sets the ranking.	Z-Image TurboAlibaba (Tongyi)	`Q4 GGUF`	6.1 GB	19.0 min	Very slowtight fit · encoder unloaded after use
76Great quality (A) · 76 of 100: how capable the model is at Q4 GGUF, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking.	Z-ImageAlibaba (Tongyi)	`Q4 GGUF`	6.1 GB	2.2 h	Very slowtight fit · encoder unloaded after use
87Excellent quality (A+) · 87 of 100: how capable the model is at Q8 GGUF, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking.	ERNIE-ImageBaidu	`Q8 GGUF`	5.4 GB	3.3 h	Very slowCPU offload

Best Video models for the Rockchip RK3566 / RK3568 (8 GB)

Quality	Model	Quant	Memory	Speed	Runs
35Basic quality (D) · 35 of 100: how capable the model is at FP16, whatever it runs on. Match score 58/100 (quality, fit, speed and memory use) sets the ranking.	AnimateDiff v3 (SD 1.5)Shanghai AI Lab	`FP16`	4.7 GB	2.0 h	Very slow
49Fair quality (C) · 49 of 100: how capable the model is at FP8, whatever it runs on. Match score 57/100 (quality, fit, speed and memory use) sets the ranking.	LTX-Video 2BLightricks	`FP8`	6.1 GB	2.4 h	Very slowtight fit · encoder unloaded after use
53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	CogVideoX 5BZhipu AI	`FP8`	5.9 GB	21.2 h	Very slowtight fit · CPU offload
53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	CogVideoX 5B I2VZhipu AI	`FP8`	6.1 GB	21.2 h	Very slowtight fit · CPU offload
53Fair quality (C) · 53 of 100: how capable the model is at FP8, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.	Wan 2.1 T2V 1.3BAlibaba	`FP8`	6.0 GB	28.2 h	Very slowtight fit · CPU offload

Best Decision models for the Rockchip RK3566 / RK3568 (8 GB)

Quality	Model	Quant	Memory	Speed	Runs
89Excellent quality (A+) · 89 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 80/100 (quality, fit, speed and memory use) sets the ranking.	Qwen3 Reranker 4BAlibaba	`Q6_K`	4.7 GB	172.9 s	Very slow
86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking.	Jina Reranker v3.5Jina AI	`Q8_0`	2.0 GB	25.9 s	Very slow
87Excellent quality (A+) · 87 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking.	zerank-2ZeroEntropy	`Q6_K`	4.7 GB	172.9 s	Very slow
86Excellent quality (A+) · 86 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 79/100 (quality, fit, speed and memory use) sets the ranking.	Shieldstral 1.0 3BMistral AI	`Q8_0`	5.4 GB	166.0 s	Very slow
84Great quality (A) · 84 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 78/100 (quality, fit, speed and memory use) sets the ranking.	Jina Reranker v3Jina AI	`Q8_0`	2.0 GB	25.9 s	Very slow
83Great quality (A) · 83 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.	Intern-Decision 4BShanghai AI Lab	`Q6_K`	4.9 GB	98.4 s	Very slow
78Great quality (A) · 78 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.	OpenAI Privacy FilterOpenAI	`Q8_0`	2.4 GB	1.2 s	Slow
82Great quality (A) · 82 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.	mxbai-rerank-large v2Mixedbread	`Q8_0`	2.8 GB	66.4 s	Very slow

Best LLMs by task on this hardware Open in Model Finder

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.