M5 Max (128 GB)
17 records across 10 models (quants: 4bit, 8bit, Q4_K - Medium, Q6_K).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M5 Max (128 GB) | Gemma 3 27B | Q6_K | llama.cpp | 20 | 391 | – | – | 8192 | – | sourced | – | source |
| M5 Max (128 GB) | Gemma 4 26B-A4B | Q4_K - Medium | MLX | 50 | – | – | 0.5 | – | – | sourced | – | source |
| M5 Max (128 GB) | Gemma-4-31B | Q4_K - Medium | MLX | 26 | – | – | 0.7 | – | – | sourced | – | source |
| M5 Max (128 GB) | pplx-decider-v1.1-27b | 4bit | MLX | – | 975 | – | – | – | 2026-10-07 | community | – | source |
| M5 Max (128 GB) | pplx-decider-v1.1-27b | 4bit | MLX | – | 730 | – | – | – | 2026-10-07 | community | – | source |
| M5 Max (128 GB) | Qwen3-Coder-Next | 8bit | MLX | 79.296 | 754.927 | – | – | 4105 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3-Coder-Next | 8bit | MLX | 74.293 | 1802.144 | – | – | 16393 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3-Coder-Next | 8bit | MLX | 68.624 | 1887.158 | – | – | 32777 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3-Coder-Next | 8bit | MLX | 48.212 | 1432.73 | – | – | 65545 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.5-122B-A10B | 4bit | MLX | 65.853 | 881.466 | – | – | 4106 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.5-122B-A10B | 4bit | MLX | 60.639 | 1239.734 | – | – | 16394 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.5-122B-A10B | 4bit | MLX | 54.923 | 1067.824 | – | – | 32778 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.5-27B | 4bit | MLX | 31.6 | – | – | – | – | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.5-27B | Q6_K | llama.cpp | 16.5 | – | – | – | 8192 | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.5-35B-A3B | Q4_K - Medium | Ollama | 48 | – | – | 0.9 | – | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.5-397B-A17B | 4bit | flash-moe | 12.99 | – | – | – | – | – | sourced | – | source |
| M5 Max (128 GB) | Qwen3.6-35B-A3B | Q4_K - Medium | MLX | 55 | – | – | 0.6 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085