M1 Max (32-core GPU, 32 GB)
3 records across 3 models (quants: Q4_K - Medium).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M1 Max (32-core GPU, 32 GB) | llama-3-1-8b-instruct | Q4_K - Medium | llamafile | 35.40364952777778 | 359.8733261944444 | – | 3.4528548829722223 | – | – | sourced | – | source |
| M1 Max (32-core GPU, 32 GB) | llama-3-2-1b-instruct | Q4_K - Medium | llamafile | 125.77495338888887 | 2025.1921614074072 | – | 0.5596487252407406 | – | – | sourced | – | source |
| M1 Max (32-core GPU, 32 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 20.07296611111111 | 195.8191797407407 | – | 6.301562639796296 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085