Community submitted benchmarks for Local AI.

Every result comes from a signed report produced on community hardware. Browse by model, quantisation and chip.

Install & start · macOS / Linux

curl -LsSf https://computearena.ai/install.sh | sh -s launch

Requires curl, tar and Python 3. Review and confirm the installation; the interactive menu opens in this terminal. No sudo required.

Latest runs

Nemotron 3 Nano 30B A3BReported as nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 · nemotron_h_moebasecompute5 Oct 2026
BaseRTQ8Apple M4 Max
94.9
decode
1,638
prefill
Qwen3.6 27BReported as Qwen/Qwen3.6-27B · qwen35basecompute4 Oct 2026
BaseRTQ8Apple M4 Max
17.5
decode
222
prefill
Qwen3.8 27BReported as Qwen/Qwen3.8-27B · qwen35basecompute3 Oct 2026
BaseRTQ4Apple M4 Max
17.7
decode
221
prefill
Qwen3.6 35B A3BReported as Qwen/Qwen3.6-35B-A3B · qwen35moebasecompute2 Oct 2026
BaseRTQ8Apple M4 Max
104.6
decode
878
prefill
Qwen3.8 27BReported as Qwen/Qwen3.8-27B · qwen35basecompute2 Oct 2026
BaseRTQ4Apple M4 Max
31.5
decode
221
prefill
Qwen3 30B A3B Thinking 2507Reported as Qwen/Qwen3-30B-A3B-Thinking-2507 · qwen3_moebasecompute2 Oct 2026
BaseRTQ8Apple M4 Max
95.9
decode
1,795
prefill

How it works

Install ComputeArena

Get the ComputeArena CLI for macOS or Linux, then choose BaseRT or llama.cpp as your runtime.

Run offline

No account is required. Signed reports remain on your machine until you submit them.

Submit when ready

Leaderboard

Model × quantisation × chip

Headline rankings compare compatible PP512 prefill and TG128 decode workloads. The measured workload is shown beside every result.

Model

Runtime

Quantisation

Chip

Prefill size

Rank by

380 matching configurations

#	Model / quant	Runtime	Chip / backend			User	Updated	Share
1	Tinystories Lay8 HS512 HD8 33MReported as ivnle/tinystories-lay8-hs512-hd8-33M · llama · default-q4 Q4	BaseRT	Apple M5 Pro Metal	2,068.0 TG128	190,107 PP512	arki05	17 Sep 2026	Compare
2	Tinystories Lay8 HS512 HD8 33MReported as RichardErkhov/ivnle_-_tinystories-lay8-hs512-hd8-33M-gguf · llama Q4_0	llama.cpp	Apple M5 Pro BLAS + Metal	1,443.3 TG128	124,181 PP512	arki05	17 Sep 2026	Compare
3	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen · default-q4 Q4	BaseRT	Apple M5 Max Metal	708.3 TG128	34,136 PP512	basecompute	18 Sep 2026	Compare
4	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen Q4	BaseRT	Apple M4 Max Metal	583.3 TG128 @ 1 ctx	9,998 PP512	basecompute	19 Sep 2026	Compare
5	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen · default-q8 Q8	BaseRT	Apple M5 Max Metal	549.5 TG128	33,088 PP512	basecompute	10 Sep 2026	Compare
6	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen · default-q4 Q4	BaseRT	Apple M5 Pro Metal	516.8 TG128	20,778 PP512	arki05	18 Sep 2026	Compare
7	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q4_K_M	llama.cpp	AMD Radeon RX 7900 XT (RADV NAVI31) Vulkan	508.8 TG128	26,137 PP512	arki05	11 Sep 2026	Compare
8	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen Q4	BaseRT	Apple M3 Ultra Metal	501.7 TG128 @ 1 ctx	12,830 PP512	basecompute	19 Sep 2026	Compare
9	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q8_0	llama.cpp	AMD Radeon RX 7900 XT (RADV NAVI31) Vulkan	479.4 TG128	26,346 PP512	arki05	11 Sep 2026	Compare
10	Qwen3 0.6BReported as Qwen3-0.6B-cuda-q4, Qwen3-0.6B-cuda-q4mix, Qwen3-0.6B-cuda-q8, Qwen/Qwen3-0.6B · qwen Q4	BaseRT	NVIDIA GB10 CUDA	458.8 TG128 @ 1 ctx	52,750 PP512	basecompute	19 Sep 2026	Compare
11	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen Q8	BaseRT	Apple M4 Max Metal	455.8 TG128 @ 1 ctx	9,749 PP512	basecompute	19 Sep 2026	Compare
12	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen Q8	BaseRT	Apple M3 Ultra Metal	440.2 TG128 @ 1 ctx	12,494 PP512	basecompute	19 Sep 2026	Compare
13	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q4_K_M	llama.cpp	Apple M5 Max BLAS + Metal	439.1 TG128	24,365 PP512	basecompute	10 Sep 2026	Compare
14	Gemma 3 1B ITReported as google/gemma-3-1b-it · gemma3 · default-q4 Q4	BaseRT	Apple M5 Max Metal	415.5 TG128	23,019 PP512	lukas	18 Sep 2026	Compare
15	Llama 3.2 1B InstructReported as meta-llama/Llama-3.2-1B-Instruct · llama Q4	BaseRT	Apple M3 Ultra Metal	410.1 TG128 @ 1 ctx	9,066 PP512	basecompute	20 Sep 2026	Compare
16	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q8_0	llama.cpp	Apple M5 Max BLAS + Metal	379.0 TG128	24,983 PP512	basecompute	10 Sep 2026	Compare
17	Llama 3.2 1B InstructReported as meta-llama/Llama-3.2-1B-Instruct · llama Q4	BaseRT	Apple M4 Max Metal	371.7 TG128 @ 1 ctx	6,058 PP512	basecompute	19 Sep 2026	Compare
18	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q4_K_M	llama.cpp	AMD Radeon RX 7900 XT ROCm	368.6 TG128	23,972 PP512	arki05	11 Sep 2026	Compare
19	Gemma 3 1B ITReported as google/gemma-3-1b-it · gemma3 Q4	BaseRT	Apple M4 Max Metal	361.2 TG128 @ 1 ctx	7,448 PP512	basecompute	19 Sep 2026	Compare
20	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q4_K_M	llama.cpp	Apple M5 Pro BLAS + Metal	358.3 TG128	14,552 PP512	arki05	11 Sep 2026	Compare
21	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen · default-q8 Q8	BaseRT	Apple M5 Pro Metal	350.2 TG128	20,537 PP512	arki05	11 Sep 2026	Compare
22	Gemma 3 1B ITReported as google/gemma-3-1b-it · gemma3 Q4	BaseRT	Apple M3 Ultra Metal	328.0 TG128 @ 1 ctx	9,903 PP512	basecompute	20 Sep 2026	Compare
23	Qwen3 1.7BReported as Qwen/Qwen3-1.7B · qwen Q4	BaseRT	Apple M4 Max Metal	318.4 TG128 @ 1 ctx	4,146 PP512	basecompute	19 Sep 2026	Compare
24	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q8_0	llama.cpp	AMD Radeon RX 7900 XT ROCm	313.7 TG128	24,691 PP512	arki05	11 Sep 2026	Compare
25	Qwen3 1.7BReported as Qwen/Qwen3-1.7B · qwen Q4	BaseRT	Apple M3 Ultra Metal	307.4 TG128 @ 1 ctx	5,420 PP512	basecompute	20 Sep 2026	Compare
26	Qwen3.5 2BReported as Qwen/Qwen3.5-2B · qwen35 Q4	BaseRT	Apple M3 Ultra Metal	306.7 TG128 @ 1 ctx	1,759 PP512	basecompute	20 Sep 2026	Compare
27	Qwen3.5 2BReported as Qwen/Qwen3.5-2B · qwen35 Q4	BaseRT	Apple M4 Max Metal	301.7 TG128 @ 1 ctx	1,718 PP512	basecompute	19 Sep 2026	Compare
28	Gemma 3 1B ITReported as gemma-3-1b-it-cuda-q4mix, gemma-3-1b-it-cuda-q8, google/gemma-3-1b-it · gemma3 Q4	BaseRT	NVIDIA GB10 CUDA	295.9 TG128 @ 1 ctx	33,841 PP512	basecompute	19 Sep 2026	Compare
29	Qwen3.5 2B BaseReported as Qwen/Qwen3.5-2B-Base · qwen35 Q4	BaseRT	Apple M4 Max Metal	295.1 TG128 @ 1 ctx	1,713 PP512	basecompute	19 Sep 2026	Compare
30	Qwen3.5 2B BaseReported as Qwen/Qwen3.5-2B-Base · qwen35 Q4	BaseRT	Apple M3 Ultra Metal	293.8 TG128 @ 1 ctx	1,756 PP512	basecompute	20 Sep 2026	Compare
31	Gemma 3 1B ITReported as google/gemma-3-1b-it · gemma3 · default-q4 Q4	BaseRT	Apple M5 Pro Metal	293.1 TG128	15,066 PP512	arki05	11 Sep 2026	Compare
32	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen Q8	BaseRT	Apple M4 Pro Metal	284.7 TG128 @ 1 ctx	4,787 PP512	basecompute	19 Sep 2026	Compare
33	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen3 Q8_0	llama.cpp	Apple M5 Pro BLAS + Metal	283.0 TG128	14,942 PP512	arki05	11 Sep 2026	Compare
34	Gemma 3 1B ITReported as google/gemma-3-1b-it · gemma3 Q8	BaseRT	Apple M4 Max Metal	281.0 TG128 @ 1 ctx	7,527 PP512	basecompute	19 Sep 2026	Compare
35	Qwen3 0.6BReported as Qwen/Qwen3-0.6B · qwen · default-q4 Q4	BaseRT	Apple M1 Pro Metal	268.8 TG128 @ 1 ctx	2,847 PP512	isu	19 Sep 2026	Compare
36	Llama 3.2 1B InstructReported as Llama-3.2-1B-Instruct-cuda-q4mix, Llama-3.2-1B-Instruct-cuda-q8, meta-llama/Llama-3.2-1B-Instruct · llama Q4	BaseRT	NVIDIA GB10 CUDA	261.3 TG128 @ 1 ctx	36,647 PP512	basecompute	19 Sep 2026	Compare
37	Gemma 3 1B ITReported as google/gemma-3-1b-it · gemma3 Q8	BaseRT	Apple M3 Ultra Metal	257.8 TG128 @ 1 ctx	9,953 PP512	basecompute	20 Sep 2026	Compare
38	Llama 3.2 1B InstructReported as meta-llama/Llama-3.2-1B-Instruct · llama Q4	BaseRT	Apple M1 Max Metal	242.1 TG128 @ 1 ctx	3,372 PP512	basecompute	20 Sep 2026	Compare
39	Qwen3 1.7BReported as Qwen/Qwen3-1.7B · qwen Q8	BaseRT	Apple M3 Ultra Metal	238.6 TG128 @ 1 ctx	5,698 PP512	basecompute	20 Sep 2026	Compare
40	Qwen3 1.7BReported as Qwen/Qwen3-1.7B · qwen · default-q4 Q4	BaseRT	Apple M5 Pro Metal	235.4 TG128	8,040 PP512	arki05	11 Sep 2026	Compare