M4 Pro (20-core GPU, 64 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 (20-core GPU, 64 GB) | llama-3-1-8b-instruct | Q4_K - Medium | llamafile | 32.90144462962963 | 349.4138904444444 | – | 3.563006918222223 | – | – | sourced | – | source |
| M4 Pro (20-core GPU, 64 GB) | llama-3-2-1b-instruct | Q4_K - Medium | llamafile | 118.5574338888889 | 2145.125925888889 | – | 0.5718198889722222 | – | – | sourced | – | source |
| M4 Pro (20-core GPU, 64 GB) | qwen-2-5-14b-instruct | Q4_K - Medium | llamafile | 18.008562833333333 | 182.97843991666667 | – | 6.9151872395 | – | – | sourced | – | source |
All rows cite their source. Provenance: community 40018 sourced 1085