Apple M4 Pro (16-core GPU) × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on Apple M4 Pro (16-core GPU)
112best gen tok/s
1,889best prompt tok/s
4,352max context tested
10results
Measured · all numbers below come from the linked sources
llama.cpp
| Quant | Gen tok/s | Prompt tok/s | Source | Context tested | TTFT | Build |
|---|---|---|---|---|---|---|
| Q4_K - Medium | 112 | 1,859 | LocalScore ↗ | 512+128 | 670 ms | llamafile 0.9.2 |
| Q4_K - Medium | 111 | 1,889 | LocalScore ↗ | 512+128 | 663 ms | llamafile 0.9.2 |
| Q4_K - Medium | 111 | 1,774 | LocalScore ↗ | 512+128 | 680 ms | llamafile 0.9.2 |
| Q4_K - Medium | 111 | 1,782 | LocalScore ↗ | 4,352 | 678 ms | llamafile 0.9.2 |
| Q4_K - Medium | 111 | 1,735 | LocalScore ↗ | 512+128 | 683 ms | llamafile 0.9.2 |
| Q4_K - Medium | 111 | 1,846 | LocalScore ↗ | 512+128 | 669 ms | llamafile 0.9.2 |
| Q4_K - Medium | 111 | 1,736 | LocalScore ↗ | 512+128 | 684 ms | llamafile 0.9.2 |
| Q4_K - Medium | 110 | 1,780 | LocalScore ↗ | 4,352 | 678 ms | llamafile 0.9.2 |
| Q4_K - Medium | 110 | 1,806 | LocalScore ↗ | 4,352 | 672 ms | llamafile 0.9.2 |
| Q4_K - Medium | 84.4 | 1,417 | LocalScore ↗ | 4,352 | 939 ms | llamafile 0.9.2 |
Ollama
No measured Ollama results yet. Submit yours →
MLX
No measured MLX results yet. Submit yours →
"Context tested" is the largest prompt+generation length actually benchmarked (tokens), not a memory-fit estimate.
What else can Apple M4 Pro (16-core GPU) run? →
Includes results from LocalScore, a Mozilla Builders project by cjpais, mirrored in the cn0303/local-llm-speed-benchmarks dataset. LocalScore runs use llamafile, which is built on llama.cpp.