M2 (10-core GPU, 16 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M2 (10-core GPU, 16 GB) | llama-3-1-8b-instruct | Q4_K - Medium | llamafile | 14.68716522222222 | 140.73035655555557 | – | 9.125759541666666 | – | – | sourced | – | source |
| M2 (10-core GPU, 16 GB) | llama-3-2-1b-instruct | Q4_K - Medium | llamafile | 55.565495822222225 | 801.8366849333333 | – | 1.589305650066667 | – | – | sourced | – | source |
| M2 (10-core GPU, 16 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 8.076025111111111 | 74.75765844444444 | – | 17.296274291666677 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085