Apple M2 Pro (16-core GPU) × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on Apple M2 Pro (16-core GPU)
91.5best gen tok/s
1,347best prompt tok/s
4,352max context tested
5results
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 | 91.5 | 1,341 | LocalScore ↗ | 512+128 | 899 ms | llamafile 0.9.2 |
| Q4_K - Medium | 91.5 | 1,281 | LocalScore ↗ | 512+128 | 914 ms | llamafile 0.9.2 |
| Q4_K - Medium | 91.0 | 1,347 | LocalScore ↗ | 4,352 | 900 ms | llamafile 0.9.2 |
| Q4_K - Medium | 90.9 | 1,305 | LocalScore ↗ | 4,352 | 908 ms | llamafile 0.9.2 |
| Q4_K - Medium | 90.8 | 1,316 | LocalScore ↗ | 512+128 | 904 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 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.