Open-weight models that run on CPU only · 16 GB

16 GB DDR5 dual channel · 89.6 GB/s
Usable 10.0 GB

Search Precision Context Concurrent requests Minimum speed Architecture Type Maker Sort

Fit	Model	Parameters / active	Precision	Memory	Est. speed	Longest context	Runs as
Runs well	Ling-3.0-tiny Ant Group	7.9B / 1.3B	Q8_0maker’s file	8.5 / 10.0 GB	28–42 tok/s	63K	System memory
Runs well	Gemma 4 E2B Google	5.1B	Q4_K_M	3.5 / 10.0 GB	14–21 tok/s	125K	System memory
Runs well	Qwen3.5-4B Alibaba Qwen	4.7B	Q5_K_M	3.8 / 10.0 GB	13–19 tok/s	205K	System memory
Runs well	Qwen3-VL-4B-Instruct Alibaba Qwen	4.4B	IQ4_XS	3.8 / 10.0 GB	13–19 tok/s	52K	System memory
Runs well	Qwen3-4B-Instruct-2507 Alibaba Qwen	4B	Q4_0	3.7 / 10.0 GB	13–19 tok/s	52K	System memory
Runs well	Qwen3-4B Alibaba Qwen	4B	Q4_0	3.7 / 10.0 GB	13–19 tok/s	128K	System memory
Runs well	DeepSeek-OCR DeepSeek	3.3B	BF16 (native)maker’s file	7.2 / 10.0 GB	16–24 tok/s	8K	System memory
Runs well	Granite-4.0-H-Micro IBM	3.2B	Q8_0maker’s file	3.7 / 10.0 GB	13–19 tok/s	128K	System memory
Runs well	Qwen3-1.7B Alibaba Qwen	2B	Q8_0maker’s file	3.1 / 10.0 GB	16–24 tok/s	40K	System memory
Runs well	DeepSeek-R1-Distill-Qwen-1.5B DeepSeek	1.8B	Q8_0	2.5 / 10.0 GB	21–32 tok/s	125K	System memory
Runs well	GLM-OCR Zhipu AI	1.3B	BF16 (native)maker’s file	3.5 / 10.0 GB	14–21 tok/s	112K	System memory
Runs well	Qwen3-0.6B Alibaba Qwen	0.8B	BF16 (native)maker’s file	2.8 / 10.0 GB	18–28 tok/s	40K	System memory
Runs well	EmbeddingGemma 2 Google	0.7B	BF16 (native)maker’s file	2.0 / 10.0 GB	28–43 tok/s	256K	System memory
Runs well	Qwen3-Embedding-0.6B Alibaba Qwen	0.6B	BF16 (native)maker’s file	2.5 / 10.0 GB	21–32 tok/s	32K	System memory
Runs well	bge-reranker-v2-m3 BAAI	0.6B	FP32 (native)maker’s file	3.4 / 10.0 GB	15–22 tok/s	8K	System memory
Runs well	bge-m3-unsupervised BAAI	0.6B	FP32 (native)maker’s file	3.4 / 10.0 GB	15–22 tok/s	8K	System memory
Runs well	bge-large-en-v1.5 BAAI	0.3B	FP32 (native)maker’s file	1.8 / 10.0 GB	32–48 tok/s	512	System memory
Runs well	bge-reranker-base BAAI	0.3B	FP32 (native)maker’s file	1.6 / 10.0 GB	40–59 tok/s	514	System memory
Runs well	bge-base-en-v1.5 BAAI	0.1B	FP32 (native)maker’s file	0.9 / 10.0 GB	98–147 tok/s	512	System memory
Runs well	bge-small-en-v1.5 BAAI	0B	FP32 (native)maker’s file	0.6 / 10.0 GB	313–470 tok/s	512	System memory
Runs, slowly	Qwen2.5-14B-Instruct-1M Alibaba Qwen	14.8B	Q4_0	9.7 / 10.0 GB	5–7 tok/s	9K	System memory
Runs, slowly	Qwen2.5-14B-Instruct Alibaba Qwen	14.8B	Q4_0maker’s file	9.9 / 10.0 GB	4–7 tok/s	8K	System memory
Runs, slowly	DeepSeek-R1-Distill-Qwen-14B DeepSeek	14.8B	Q4_0	9.7 / 10.0 GB	5–7 tok/s	9K	System memory
Runs, slowly	Qwen3-14B Alibaba Qwen	14.8B	Q4_0	9.5 / 10.0 GB	5–7 tok/s	11K	System memory
Runs, slowly	Gemma 3 12B IT Google	12.2B	Q4_0	9.7 / 10.0 GB	5–7 tok/s	8K	System memory
Runs, slowly	Qwen2.5-Omni-7B Alibaba Qwen	10.7B	Q5_K_M	8.0 / 10.0 GB	6–8 tok/s	32K	System memory
Runs, slowly	GLM-4.6V-Flash Zhipu AI	10.3B	Q6_K	8.7 / 10.0 GB	5–8 tok/s	41K	System memory
Runs, slowly	Qwen3.5-9B Alibaba Qwen	9.7B	Q6_K	8.1 / 10.0 GB	5–8 tok/s	68K	System memory
Runs, slowly	Qianfan-VL-8B Baidu	8.8B	Q6_K	8.2 / 10.0 GB	5–8 tok/s	22K	System memory
Runs, slowly	Granite Switch 4.1 8B IBM	8.8B	Q5_K_Mmaker’s file	8.5 / 10.0 GB	5–8 tok/s	17K	System memory
Runs, slowly	Qwen3-VL-8B-Instruct Alibaba Qwen	8.8B	Q6_K	8.3 / 10.0 GB	5–8 tok/s	20K	System memory
Runs, slowly	Qwen3-VL-8B-Thinking Alibaba Qwen	8.8B	Q6_K	8.3 / 10.0 GB	5–8 tok/s	20K	System memory
Runs, slowly	pplx-embed-v2-context-9b Perplexity	8.4B	FP8	8.6 / 10.0 GB	5–8 tok/s	53K	System memory
Runs, slowly	Qwen2.5-VL-7B-Instruct Alibaba Qwen	8.3B	FP8	8.7 / 10.0 GB	5–8 tok/s	32K	System memory
Runs, slowly	Qwen3-8B Alibaba Qwen	8.2B	Q6_Kmaker’s file	7.9 / 10.0 GB	6–8 tok/s	23K	System memory
Runs, slowly	DeepSeek-R1-Distill-Llama-8B DeepSeek	8B	FP8	9.0 / 10.0 GB	5–7 tok/s	16K	System memory
Runs, slowly	Gemma 4 E4B Google	8B	Q8_0	8.6 / 10.0 GB	5–8 tok/s	113K	System memory
Runs, slowly	Qwen2.5-7B-Instruct-1M Alibaba Qwen	7.6B	Q8_0	8.5 / 10.0 GB	5–8 tok/s	36K	System memory
Runs, slowly	Qwen2.5-7B-Instruct Alibaba Qwen	7.6B	Q8_0maker’s file	8.5 / 10.0 GB	5–8 tok/s	32K	System memory
Runs, slowly	DeepSeek-R1-Distill-Qwen-7B DeepSeek	7.6B	Q8_0	8.5 / 10.0 GB	5–8 tok/s	32K	System memory
Runs, slowly	Mistral-7B-Instruct-v0.1 Mistral AI	7.2B	Q8_0	8.7 / 10.0 GB	5–8 tok/s	18K	System memory
Runs, slowly	WeVisOne-VL-4B Tencent	4.8B	Q8_0	6.4 / 10.0 GB	7–11 tok/s	33K	System memory
Runs, slowly	Qianfan-OCR Baidu	4.7B	Q8_0	6.3 / 10.0 GB	7–11 tok/s	31K	System memory

Estimates from the makers’ Hugging Face repositories (as of 2026-11-06); filtered views aren’t indexed. How it’s estimated