M2
72 records across 15 models (quants: 1bit, 4bit, 8bit, F16, Q4_0, Q8_0, oQ4e).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M2 | Bonsai-4B | 1bit | oMLX | 43.7 | 259.9 | – | – | 1024 | 2026-10-14 | community | – | source |
| M2 | Gemma 4 E2B | 4bit | oMLX | 50.9 | 1260.0 | – | – | 4096 | 2026-10-07 | community | – | source |
| M2 | Gemma 4 E2B | 4bit | oMLX | 43.6 | 1295.0 | – | – | 16384 | 2026-10-07 | community | – | source |
| M2 | Gemma 4 E4B | 4bit | oMLX | 29.1 | 393.9 | – | – | 1024 | 2026-10-14 | community | – | source |
| M2 | gemma-4-E2B-it | oQ4e | oMLX | 51.1 | 1168.0 | – | – | 4096 | 2026-10-10 | community | – | source |
| M2 | gemma-4-E2B-it | oQ4e | oMLX | 49.3 | 1150.0 | – | – | 1024 | 2026-10-10 | community | – | source |
| M2 | gemma-4-E2B-it | oQ4e | oMLX | 45.9 | 1162.0 | – | – | 1024 | 2026-10-10 | community | – | source |
| M2 | gemma-4-E2B-it | oQ4e | oMLX | 44.8 | 1137.0 | – | – | 8192 | 2026-10-10 | community | – | source |
| M2 | gemma-4-E2B-it | oQ4e | oMLX | 38.5 | 1057.0 | – | – | 16384 | 2026-10-10 | community | – | source |
| M2 | gemma-4-E2B-it | oQ4e | oMLX | 22.2 | 596.5 | – | – | 65536 | 2026-10-10 | community | – | source |
| M2 | gemma-4-E2B-it | oQ4e | oMLX | 21.6 | 751.5 | – | – | 32768 | 2026-10-10 | community | – | source |
| M2 | gemma-4-E2B-it-Uncensored-MAX | 4bit | oMLX | 50.1 | 769.2 | – | – | 1024 | 2026-10-15 | community | – | source |
| M2 | granite-4.2-3b | 4bit | oMLX | 11.8 | 68.5 | – | – | 1024 | 2026-10-15 | community | – | source |
| M2 | granite-4.2-3b | 4bit | oMLX | 9.8 | 64.7 | – | – | 4096 | 2026-10-15 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 142.13 | 1411.25 | – | – | tg128 | 2025-11-19 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 129.58 | 1391.39 | – | – | tg128 | 2025-11-19 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 109.21 | 792.74 | – | – | tg128 | 2025-10-18 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 105.06 | 796.29 | – | – | tg128 | 2025-10-18 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 66.48 | 685.51 | – | – | tg128 | 2025-11-26 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 66.32 | 683.61 | – | – | tg128 | 2025-11-28 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 63.16 | 678.14 | – | – | tg128 | 2025-11-26 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 63.1 | 675.02 | – | – | tg128 | 2025-11-28 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 62.75 | 53.16 | – | – | tg128 | 2025-11-26 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 58.47 | 53.1 | – | – | tg128 | 2025-11-26 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 13.5 | 50.79 | – | – | tg128 | 2025-11-07 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 12.25 | 46.81 | – | – | tg128 | 2025-11-07 | community | – | source |
| M2 | llama 7B | Q4_0 | llama.cpp (Vulkan) | 11.07 | 38.67 | – | – | tg128 | 2025-01-08 | community | – | source |
| M2 | llama 7B mostly | Q4_0 | llama.cpp (Metal) | 64.55 | 669.51 | – | – | tg128 | 2023-11-25 | community | – | source |
| M2 | llama 7B mostly | Q4_0 | llama.cpp (Metal) | 63.29 | 664.07 | – | – | tg128 | 2023-11-25 | community | – | source |
| M2 | llama 7B mostly | Q4_0 | llama.cpp (Metal) | 62.98 | 668.28 | – | – | tg128 | 2023-11-26 | community | – | source |
| M2 | llama 7B mostly | Q8_0 | llama.cpp (Metal) | 41.79 | 674.5 | – | – | tg128 | 2023-11-25 | community | – | source |
| M2 | llama 7B mostly | Q8_0 | llama.cpp (Metal) | 40.85 | 671.33 | – | – | tg128 | 2023-11-25 | community | – | source |
| M2 | llama 7B mostly | Q8_0 | llama.cpp (Metal) | 40.67 | 674.37 | – | – | tg128 | 2023-11-26 | community | – | source |
| M2 | llama 7B mostly | F16 | llama.cpp (Metal) | 24.54 | 747.99 | – | – | tg128 | 2023-11-26 | community | – | source |
| M2 | llama 7B mostly | F16 | llama.cpp (Metal) | 24.31 | 754.39 | – | – | tg128 | 2023-11-25 | community | – | source |
| M2 | llama 7B mostly | Q4_0 | llama.cpp (Metal) | 21.91 | 179.57 | – | – | tg128 | 2023-11-22 | community | – | source |
| M2 | llama 7B mostly | Q4_0 | llama.cpp (Metal) | 21.7 | 145.91 | – | – | tg128 | 2024-01-07 | community | – | source |
| M2 | llama 7B mostly | Q8_0 | llama.cpp (Metal) | 12.21 | 181.4 | – | – | tg128 | 2023-11-22 | community | – | source |
| M2 | llama 7B mostly | Q8_0 | llama.cpp (Metal) | 12.18 | 147.27 | – | – | tg128 | 2024-01-07 | community | – | source |
| M2 | llama 7B mostly | Q4_0 | llama.cpp (Metal) | 8.61 | 85.24 | – | – | tg128 | 2023-12-05 | community | contradiction | source |
| M2 | llama 7B mostly | F16 | llama.cpp (Metal) | 6.72 | 201.34 | – | – | tg128 | 2023-11-22 | community | – | source |
| M2 | llama 7B mostly | Q8_0 | llama.cpp (Metal) | 5.31 | 76.9 | – | – | tg128 | 2023-12-05 | community | contradiction | source |
| M2 | Qwen2.5-Coder-3B-Instruct | 4bit | oMLX | 15.8 | 212.2 | – | – | 1024 | 2026-10-15 | community | – | source |
| M2 | Qwen2.5-Coder-3B-Instruct | 4bit | oMLX | 15.3 | 129.7 | – | – | 1024 | 2026-10-15 | community | – | source |
| M2 | Qwen2.5-Coder-3B-Instruct | 4bit | oMLX | 13.6 | 83.6 | – | – | 4096 | 2026-10-15 | community | – | source |
| M2 | Qwen2.5-Coder-3B-Instruct | 4bit | oMLX | 13.5 | 81.7 | – | – | 4096 | 2026-10-15 | community | – | source |
| M2 | Qwen3 14B | 4bit | oMLX | 10.1 | 70.7 | – | – | 4096 | 2026-10-08 | community | – | source |
| M2 | Qwen3 14B | 4bit | oMLX | 9.2 | 67.1 | – | – | 8192 | 2026-10-08 | community | – | source |
| M2 | Qwen3 8B | 4bit | oMLX | 16.7 | 128.9 | – | – | 4096 | 2026-10-12 | community | – | source |
| M2 | Qwen3 8B | 4bit | oMLX | 11.9 | 105.0 | – | – | 16384 | 2026-10-12 | community | – | source |
| M2 | Qwen3-0.6B | 8bit | oMLX | 86.8 | 1749.0 | – | – | 1024 | 2026-10-06 | community | – | source |
| M2 | Qwen3-0.6B | 8bit | oMLX | 60.7 | 1317.0 | – | – | 4096 | 2026-10-06 | community | – | source |
| M2 | Qwen3.5-4B-Claude-4.6-Opus-Reasoning-Distilled | 4bit | oMLX | 32.5 | 251.8 | – | – | 1024 | 2026-10-12 | community | – | source |
| M2 | Qwen3.5-4B-Claude-4.6-Opus-Reasoning-Distilled | 4bit | oMLX | 30.1 | 235.5 | – | – | 4096 | 2026-10-12 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.8 | 110.4 | – | – | 1024 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.8 | 110.4 | – | – | 1024 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.8 | 110.4 | – | – | 1024 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.6 | 105.8 | – | – | 4096 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.6 | 105.8 | – | – | 4096 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.6 | 105.8 | – | – | 4096 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.5 | 86.4 | – | – | 4096 | 2026-10-12 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.5 | 109.9 | – | – | 1024 | 2026-10-17 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 18.3 | 105.5 | – | – | 4096 | 2026-10-17 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 17.8 | 101.5 | – | – | 8192 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 17.8 | 101.5 | – | – | 8192 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 17.8 | 101.5 | – | – | 8192 | 2026-10-11 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 17.6 | 104.7 | – | – | 4096 | 2026-10-14 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 17.4 | 85.4 | – | – | 8192 | 2026-10-12 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 14.6 | 43.8 | – | – | 16384 | 2026-10-12 | community | – | source |
| M2 | Qwen3.5-9B | 4bit | oMLX | 10.2 | 77.9 | – | – | 16384 | 2026-10-14 | community | contradiction | source |
| M2 | Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2 | 4bit | oMLX | 18.6 | 110.1 | – | – | 1024 | 2026-10-09 | community | – | source |
| M2 | Qwen3.5-9B-Claude-4.6-Opus-Reasoning-Distilled-v2 | 4bit | oMLX | 17.9 | 105.7 | – | – | 4096 | 2026-10-09 | community | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085