M3 Pro

163 records across 28 models (quants: 2bit, 4bit, 6bit, 8bit, F16, MXFP4-Q8, OptiQ 4bit, OptiQ-4bit, Q4_0, Q4_K_M, Q6_K, Q8_0, oQ2.5e, oQ3e, oQ4e, oQ8e).

hardware	model	quant	backend	tok/s	pp tok/s	W	ttft s	ctx/tg	date	class	flags	source
M3 Pro	Devstral-Small-2-24B-Instruct-2512	6bit	oMLX	6.4	111.4	–	–	4096	2026-10-09	community	–	source
M3 Pro	Devstral-Small-2-24B-Instruct-2512	6bit	oMLX	5.7	103.2	–	–	16384	2026-10-09	community	–	source
M3 Pro	Gemma 4 12B	8bit	oMLX	10.2	213.8	–	–	1024	2026-10-14	community	–	source
M3 Pro	Gemma 4 12B	8bit	oMLX	10.2	212.7	–	–	1024	2026-10-14	community	–	source
M3 Pro	Gemma 4 12B	8bit	oMLX	10.1	216.5	–	–	1024	2026-10-14	community	–	source
M3 Pro	Gemma 4 12B	8bit	oMLX	10.1	210.9	–	–	1024	2026-10-14	community	–	source
M3 Pro	Gemma 4 12B	8bit	oMLX	10.1	215.2	–	–	4096	2026-10-14	community	–	source
M3 Pro	Gemma 4 E4B	4bit	oMLX	43.2	996.1	–	–	1024	2026-10-14	community	–	source
M3 Pro	Gemma 4 E4B	4bit	oMLX	35.8	774.2	–	–	1024	2026-10-14	community	contradiction	source
M3 Pro	gemma-4-12b-coder-fable5-composer2.5	4bit	oMLX	18.0	170.1	–	–	4096	2026-10-13	community	–	source
M3 Pro	gemma-4-12b-coder-fable5-composer2.5	4bit	oMLX	17.3	159.3	–	–	16384	2026-10-13	community	–	source
M3 Pro	gemma-4-12b-it	oQ4e	oMLX	26.6	207.4	–	–	4096	2026-10-15	community	–	source
M3 Pro	gemma-4-12b-it	4bit	oMLX	18.1	217.0	–	–	1024	2026-10-14	community	–	source
M3 Pro	gemma-4-12b-it	4bit	oMLX	17.9	217.0	–	–	4096	2026-10-14	community	–	source
M3 Pro	gemma-4-12b-it	4bit	oMLX	17.3	221.1	–	–	1024	2026-10-14	community	–	source
M3 Pro	gemma-4-12b-it	oQ4e	oMLX	17.2	210.8	–	–	1024	2026-10-15	community	–	source
M3 Pro	gemma-4-12b-it	oQ8e	oMLX	14.3	212.9	–	–	4096	2026-10-15	community	–	source
M3 Pro	gemma-4-12b-it	oQ8e	oMLX	12.9	214.0	–	–	1024	2026-10-15	community	–	source
M3 Pro	gemma-4-12b-it	oQ4e	oMLX	11.6	196.5	–	–	16384	2026-10-15	community	–	source
M3 Pro	gemma-4-12b-it	oQ8e	oMLX	9.8	196.6	–	–	16384	2026-10-15	community	–	source
M3 Pro	gemma-4-12B-it-qat	8bit	oMLX	17.2	213.1	–	–	1024	2026-10-12	community	–	source
M3 Pro	gemma-4-12B-it-qat	8bit	oMLX	14.2	202.1	–	–	4096	2026-10-12	community	–	source
M3 Pro	gemma-4-12B-it-qat	8bit	oMLX	10.2	211.4	–	–	1024	2026-10-12	community	contradiction	source
M3 Pro	gemma-4-12B-it-qat	8bit	oMLX	9.9	208.5	–	–	4096	2026-10-12	community	contradiction	source
M3 Pro	gemma4 26B.A4B	4bit	oMLX	47.2	453.7	–	–	16384	2026-10-12	community	–	source
M3 Pro	gemma4 26B.A4B	4bit	oMLX	40.9	571.9	–	–	4096	2026-10-09	community	–	source
M3 Pro	gemma4 26B.A4B	4bit	oMLX	37.5	534.6	–	–	16384	2026-10-09	community	contradiction	source
M3 Pro	gemma4 26B.A4B	4bit	oMLX	27.0	418.7	–	–	4096	2026-10-12	community	contradiction	source
M3 Pro	gpt-oss 20B	OptiQ-4bit	oMLX	52.8	642.4	–	–	1024	2026-10-12	community	–	source
M3 Pro	gpt-oss 20B	OptiQ-4bit	oMLX	50.8	635.6	–	–	4096	2026-10-12	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	41.2	664.2	–	–	4096	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	41.2	664.2	–	–	4096	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	36.5	584.5	–	–	16384	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	36.5	584.5	–	–	16384	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	30.8	502.3	–	–	32768	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	30.8	502.3	–	–	32768	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	OptiQ-4bit	oMLX	26.1	526.4	–	–	4096	2026-10-13	community	contradiction	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	23.1	390.2	–	–	65536	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	23.1	390.2	–	–	65536	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	MXFP4-Q8	oMLX	17.0	266.8	–	–	131072	2026-10-15	community	–	source
M3 Pro	gpt-oss 20B	OptiQ-4bit	oMLX	13.8	619.3	–	–	1024	2026-10-13	community	contradiction	source
M3 Pro	Huihui-Qwen3.5-9B	8bit	oMLX	15.2	283.1	–	–	4096	2026-10-10	community	–	source
M3 Pro	Huihui-Qwen3.5-9B	8bit	oMLX	14.5	271.7	–	–	16384	2026-10-10	community	–	source
M3 Pro	LFM2.5-24B-A2B	4bit	oMLX	71.5	917.4	–	–	1024	2026-10-14	community	–	source
M3 Pro	LFM2.5-24B-A2B	4bit	oMLX	68.2	989.9	–	–	4096	2026-10-14	community	–	source
M3 Pro	llama 7B	Q4_0	llama.cpp (Metal)	31.54	371.72	–	–	tg128	2025-02-11	community	–	source
M3 Pro	llama 7B	Q4_0	llama.cpp (Metal)	31.52	368.92	–	–	tg128	2025-02-11	community	–	source
M3 Pro	llama 7B	Q4_0	llama.cpp (Vulkan)	26.39	263.7	–	–	tg128	2025-02-11	community	–	source
M3 Pro	llama 7B	Q4_0	llama.cpp (Vulkan)	26.23	259.11	–	–	tg128	2025-02-11	community	–	source
M3 Pro	llama 7B mostly	Q4_0	llama.cpp (Metal)	30.74	341.67	–	–	tg128	2023-12-09	community	–	source
M3 Pro	llama 7B mostly	Q4_0	llama.cpp (Metal)	30.65	269.49	–	–	tg128	2023-11-25	community	–	source
M3 Pro	llama 7B mostly	Q8_0	llama.cpp (Metal)	17.53	344.66	–	–	tg128	2023-12-09	community	–	source
M3 Pro	llama 7B mostly	Q8_0	llama.cpp (Metal)	17.44	272.11	–	–	tg128	2023-11-25	community	–	source
M3 Pro	llama 7B mostly	F16	llama.cpp (Metal)	9.89	357.45	–	–	tg128	2023-12-09	community	–	source
M3 Pro	Midm-2.0-Mini	Q4_K_M	llama.cpp	44.62	–	–	–	–	2026-09	community	–	source
M3 Pro	Midm-2.0-Mini	Q8_0	llama.cpp	34.26	–	–	–	–	2026-09	community	–	source
M3 Pro	Mistral-7B-Instruct-v0.3	4bit	oMLX	26.7	349.5	–	–	4096	2026-10-09	community	–	source
M3 Pro	Mistral-7B-Instruct-v0.3	4bit	oMLX	17.7	293.6	–	–	16384	2026-10-09	community	–	source
M3 Pro	Nanbeige4.2-3B	OptiQ 4bit	oMLX	24.3	387.2	–	–	1024	2026-10-09	community	–	source
M3 Pro	Nanbeige4.2-3B	OptiQ 4bit	oMLX	21.8	352.8	–	–	4096	2026-10-09	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	4bit	oMLX	55.1	770.0	–	–	4096	2026-10-09	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	4bit	oMLX	49.9	679.4	–	–	16384	2026-10-09	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	6bit	oMLX	42.3	749.8	–	–	4096	2026-10-09	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	OptiQ-4bit	oMLX	41.9	632.7	–	–	1024	2026-10-16	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	OptiQ-4bit	oMLX	41.6	638.2	–	–	1024	2026-10-16	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	OptiQ-4bit	oMLX	40.9	666.9	–	–	4096	2026-10-16	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	OptiQ-4bit	oMLX	40.7	666.3	–	–	4096	2026-10-16	community	–	source
M3 Pro	Ornith-1.5-35B-A3B	6bit	oMLX	38.1	662.1	–	–	16384	2026-10-09	community	–	source
M3 Pro	Qwen3-4B-Instruct	Q4_K_M	llama.cpp	30.81	–	–	–	–	2026-09	community	–	source
M3 Pro	Qwen3-4B-Instruct	Q8_0	llama.cpp	24.22	–	–	–	–	2026-09	community	–	source
M3 Pro	Qwen3.5-4B-OptiQ	4bit	oMLX	35.9	525.7	–	–	4096	2026-10-09	community	–	source
M3 Pro	Qwen3.5-4B-OptiQ	4bit	oMLX	35.9	525.7	–	–	4096	2026-10-09	community	–	source
M3 Pro	Qwen3.5-4B-OptiQ	4bit	oMLX	31.2	474.7	–	–	16384	2026-10-09	community	–	source
M3 Pro	Qwen3.5-4B-OptiQ	4bit	oMLX	31.2	474.7	–	–	16384	2026-10-09	community	–	source
M3 Pro	Qwen3.5-9B	oQ4e	oMLX	46.2	339.1	–	–	16384	2026-10-09	community	–	source
M3 Pro	Qwen3.5-9B	oQ4e	oMLX	43.3	351.9	–	–	4096	2026-10-09	community	–	source
M3 Pro	Qwen3.5-9B	4bit	oMLX	28.3	363.0	–	–	1024	2026-10-14	community	–	source
M3 Pro	Qwen3.5-9B	4bit	oMLX	28.2	363.3	–	–	1024	2026-10-14	community	–	source
M3 Pro	Qwen3.5-9B	4bit	oMLX	27.9	351.7	–	–	1024	2026-10-14	community	–	source
M3 Pro	Qwen3.5-9B	4bit	oMLX	27.0	293.2	–	–	4096	2026-10-09	community	–	source
M3 Pro	Qwen3.5-9B	4bit	oMLX	26.6	373.2	–	–	8192	2026-10-14	community	–	source
M3 Pro	Qwen3.5-9B	4bit	oMLX	24.1	278.2	–	–	16384	2026-10-09	community	–	source
M3 Pro	Qwen3.5-9B	4bit	oMLX	17.2	235.9	–	–	1024	2026-10-10	community	contradiction	source
M3 Pro	Qwen3.6-27B	4bit	oMLX	8.8	102.3	–	–	1024	2026-10-14	community	–	source
M3 Pro	Qwen3.6-27B	4bit	oMLX	8.6	103.5	–	–	4096	2026-10-18	community	–	source
M3 Pro	Qwen3.6-27B	4bit	oMLX	8.3	102.0	–	–	1024	2026-10-18	community	–	source
M3 Pro	Qwen3.6-27B	4bit	oMLX	6.5	89.9	–	–	1024	2026-10-13	community	contradiction	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	58.9	681.4	–	–	1024	2026-10-13	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	58.7	538.3	–	–	1024	2026-10-16	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	58.4	678.6	–	–	1024	2026-10-18	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	58.3	720.0	–	–	1024	2026-10-16	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	57.8	710.7	–	–	1024	2026-10-14	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	57.6	707.4	–	–	1024	2026-10-14	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	57.5	630.4	–	–	4096	2026-10-16	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	57.4	773.7	–	–	4096	2026-10-18	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	57.0	788.8	–	–	4096	2026-10-13	community	–	source
M3 Pro	Qwen3.6-35B-A3B	4bit	oMLX	45.9	795.7	–	–	4096	2026-10-16	community	contradiction	source
M3 Pro	Qwen3.6-35B-A3B	OptiQ-4bit	oMLX	40.6	594.5	–	–	1024	2026-10-13	community	–	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	20.5	102.8	–	–	4096	2026-10-09	community	–	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	20.2	98.8	–	–	1024	2026-10-09	community	–	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	20.0	102.4	–	–	4096	2026-10-14	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	18.6	103.9	–	–	4096	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	18.3	103.3	–	–	4096	2026-10-09	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	18.0	99.7	–	–	1024	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	17.9	103.9	–	–	4096	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	17.4	102.2	–	–	1024	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	16.7	98.7	–	–	1024	2026-10-09	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	16.6	103.9	–	–	4096	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	16.4	103.9	–	–	4096	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	16.3	102.3	–	–	1024	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	15.9	96.7	–	–	4096	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	15.8	102.2	–	–	1024	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	15.1	95.6	–	–	1024	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	14.0	102.3	–	–	1024	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ3e	oMLX	13.8	96.1	–	–	1024	2026-10-12	community	–	source
M3 Pro	Qwen3.8-27B	oQ3e	oMLX	13.7	97.2	–	–	4096	2026-10-12	community	–	source
M3 Pro	Qwen3.8-27B	oQ2.5e	oMLX	13.7	99.6	–	–	16384	2026-10-13	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	13.7	106.5	–	–	4096	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	13.5	103.9	–	–	1024	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ3e	oMLX	12.9	92.4	–	–	16384	2026-10-12	community	–	source
M3 Pro	Qwen3.8-27B	oQ3e	oMLX	12.6	97.4	–	–	4096	2026-10-12	community	–	source
M3 Pro	Qwen3.8-27B	oQ3e	oMLX	12.4	96.4	–	–	1024	2026-10-12	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ2.5e	oMLX	11.8	99.7	–	–	1024	2026-10-13	community	–	source
M3 Pro	Qwen3.8-27B	oQ2.5e	oMLX	11.8	103.5	–	–	4096	2026-10-13	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	9.0	102.9	–	–	1024	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.9	102.9	–	–	1024	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.8	102.2	–	–	1024	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.8	102.6	–	–	1024	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.7	102.3	–	–	1024	2026-10-07	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.7	105.8	–	–	4096	2026-10-07	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.7	106.2	–	–	4096	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.7	99.3	–	–	4096	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	8.2	90.9	–	–	1024	2026-10-13	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	8.1	101.5	–	–	1024	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	7.9	97.6	–	–	1024	2026-10-12	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	7.8	104.2	–	–	4096	2026-10-14	community	contradiction	source
M3 Pro	Qwen3.8-27B	oQ4e	oMLX	7.7	100.4	–	–	4096	2026-10-12	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	7.5	94.9	–	–	32768	2026-10-17	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	5.2	39.3	–	–	16384	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	5.1	64.5	–	–	8192	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	4.6	40.7	–	–	8192	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	3.5	47.1	–	–	16384	2026-10-15	community	contradiction	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	1.7	36.2	–	–	65536	2026-10-14	community	–	source
M3 Pro	Qwen3.8-27B	4bit	oMLX	1.7	26.5	–	–	65536	2026-10-15	community	–	source
M3 Pro	Qwen3.8-27B-Uncensored	OptiQ-4bit	oMLX	7.1	81.9	–	–	4096	2026-10-18	community	–	source
M3 Pro	Qwen3.8-27B-Uncensored	OptiQ-4bit	oMLX	6.8	77.9	–	–	16384	2026-10-18	community	–	source
M3 Pro	Qwen3.8-27B-Uncensored	OptiQ-4bit	oMLX	6.7	86.3	–	–	1024	2026-10-13	community	–	source
M3 Pro	SmolLM2-135M	Q6_K	llama.cpp	249.5	–	–	–	–	2026-09	community	–	source
M3 Pro	SmolLM2-135M	Q8_0	llama.cpp	205.91	–	–	–	–	2026-09	community	–	source
M3 Pro	Spark-X2.5-4B	4bit	oMLX	33.0	452.7	–	–	8192	2026-10-16	community	–	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	11.3	86.6	–	–	1024	2026-10-09	community	–	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	11.0	85.4	–	–	1024	2026-10-09	community	–	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	11.0	86.0	–	–	4096	2026-10-09	community	–	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	10.7	81.9	–	–	1024	2026-10-16	community	–	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	10.6	84.5	–	–	4096	2026-10-09	community	–	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	10.4	81.5	–	–	1024	2026-10-15	community	–	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	9.6	83.8	–	–	4096	2026-10-16	community	contradiction	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	9.2	84.3	–	–	4096	2026-10-15	community	contradiction	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	7.0	86.7	–	–	1024	2026-10-09	community	contradiction	source
M3 Pro	Ternary-Bonsai-2-27B	2bit	oMLX	6.3	86.2	–	–	4096	2026-10-09	community	contradiction	source
M3 Pro	Ternary-Bonsai-27B	2bit	oMLX	17.2	102.8	–	–	1024	2026-10-17	community	–	source
M3 Pro	Tiel-Coder-35B-A3B	oQ4e	oMLX	48.8	535.9	–	–	32768	2026-10-18	community	–	source
M3 Pro	Tiel-Coder-35B-A3B	oQ4e	oMLX	30.3	377.6	–	–	65536	2026-10-18	community	–	source

All rows cite their source. Provenance: community 40018 sourced 1085

← all hardware