Apple M1 Max (32-core GPU) × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on Apple M1 Max (32-core GPU)
128best gen tok/s
2,122best prompt tok/s
4,352max context tested
12results
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 | 128 | 2,122 | LocalScore ↗ | 512+128 | 539 ms | llamafile 0.9.2 |
| Q4_K - Medium | 126 | 1,971 | LocalScore ↗ | 512+128 | 572 ms | llamafile 0.9.2 |
| Q4_K - Medium | 126 | 2,065 | LocalScore ↗ | 512+128 | 556 ms | llamafile 0.9.2 |
| Q4_K - Medium | 126 | 2,043 | LocalScore ↗ | 512+128 | 555 ms | llamafile 0.9.2 |
| Q4_K - Medium | 125 | 2,012 | LocalScore ↗ | 4,352 | 559 ms | llamafile 0.9.2 |
| Q4_K - Medium | 124 | 2,011 | LocalScore ↗ | 512+128 | 557 ms | llamafile 0.9.2 |
| Q4_K - Medium | 124 | 2,060 | LocalScore ↗ | 512+128 | 552 ms | llamafile 0.9.2 |
| Q4_K - Medium | 124 | 2,022 | LocalScore ↗ | 512+128 | 559 ms | llamafile 0.9.2 |
| Q4_K - Medium | 124 | 1,940 | LocalScore ↗ | 512+128 | 578 ms | llamafile 0.9.2 |
| Q4_K - Medium | 123 | 1,961 | LocalScore ↗ | 512+128 | 575 ms | llamafile 0.9.2 |
| Q4_K - Medium | 110 | 1,845 | LocalScore ↗ | 512+128 | 627 ms | llamafile 0.9.2 |
| Q4_K - Medium | 92.8 | 1,689 | LocalScore ↗ | 4,352 | 678 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 M1 Max (32-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.