M4 Max (40-core GPU, 48 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M4 Max (40-core GPU, 48 GB) | llama-3-1-8b-instruct | Q4_K - Medium | llamafile | 55.0887822962963 | 663.3646092592592 | – | 1.7937055694814816 | – | – | sourced | – | source |
| M4 Max (40-core GPU, 48 GB) | llama-3-2-1b-instruct | Q4_K - Medium | llamafile | 178.9852262222222 | 3750.568106722222 | – | 0.30928531144444443 | – | – | sourced | – | source |
| M4 Max (40-core GPU, 48 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 30.13459711111111 | 347.03543255555553 | – | 3.4913595508333337 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085