M3 Ultra (60-core GPU, 96 GB)
1 records across 1 models (quants: Q4_K - Medium).
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M3 Ultra (60-core GPU, 96 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 34.43578599999999 | 444.62243322222224 | – | 2.672898578888889 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085