Apple M2 (10-core GPU) × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on Apple M2 (10-core GPU)
60.0best gen tok/s
846best prompt tok/s
640max context tested
11results
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 | 60.0 | 837 | LocalScore ↗ | 512+128 | 1,554 ms | llamafile 0.9.2 |
| Q4_K - Medium | 60.0 | 835 | LocalScore ↗ | 512+128 | 1,555 ms | llamafile 0.9.2 |
| Q4_K - Medium | 60.0 | 846 | LocalScore ↗ | 512+128 | 1,553 ms | llamafile 0.9.2 |
| Q4_K - Medium | 59.9 | 827 | LocalScore ↗ | 512+128 | 1,559 ms | llamafile 0.9.2 |
| Q4_K - Medium | 57.7 | 793 | LocalScore ↗ | 512+128 | 1,624 ms | llamafile 0.9.2 |
| Q4_K - Medium | 57.0 | 831 | LocalScore ↗ | 512+128 | 1,577 ms | llamafile 0.9.2 |
| Q4_K - Medium | 57.0 | 819 | LocalScore ↗ | 512+128 | 1,576 ms | llamafile 0.9.2 |
| Q4_K - Medium | 56.5 | 805 | LocalScore ↗ | 512+128 | 1,669 ms | llamafile 0.9.2 |
| Q4_K - Medium | 53.9 | 822 | LocalScore ↗ | 512+128 | 1,586 ms | llamafile 0.9.2 |
| Q4_K - Medium | 48.3 | 774 | LocalScore ↗ | 512+128 | 1,597 ms | llamafile 0.9.2 |
| Q4_K - Medium | 43.1 | 725 | LocalScore ↗ | 512+128 | 1,638 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 M2 (10-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.