M4 Pro (16-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 Pro (16-core GPU, 48 GB) | llama-3-1-8b-instruct | Q4_K - Medium | llamafile | 30.176443111111105 | 302.390073 | – | 4.143549694333333 | – | – | sourced | – | source |
| M4 Pro (16-core GPU, 48 GB) | llama-3-2-1b-instruct | Q4_K - Medium | llamafile | 111.00253472222222 | 1754.577648111111 | – | 0.6816085969999999 | – | – | sourced | – | source |
| M4 Pro (16-core GPU, 48 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 16.762545222222222 | 161.09253933333332 | – | 7.843727675888888 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085