M2 Max (38-core GPU, 96 GB)
3 records across 3 models (quants: 8bit, Q4_K - Medium).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M2 Max (38-core GPU, 96 GB) | llama-3-1-8b-instruct | Q4_K - Medium | llamafile | 46.41150744444444 | 484.057812 | – | 2.4422629954444446 | – | – | sourced | – | source |
| M2 Max (38-core GPU, 96 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 25.230792444444447 | 252.470057 | – | 4.798504768333334 | – | – | sourced | – | source |
| M2 Max (38-core GPU, 96 GB) | Qwen3.5-27B | 8bit | MLX | 12 | – | – | – | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085