Apple M1 (8-core GPU) × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on Apple M1 (8-core GPU)
41.4best gen tok/s
615best 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 | 41.4 | 612 | LocalScore ↗ | 512+128 | 2,150 ms | llamafile 0.9.2 |
| Q4_K - Medium | 41.2 | 608 | LocalScore ↗ | 512+128 | 2,181 ms | llamafile 0.9.2 |
| Q4_K - Medium | 40.9 | 601 | LocalScore ↗ | 512+128 | 2,180 ms | llamafile 0.9.2 |
| Q4_K - Medium | 40.7 | 615 | LocalScore ↗ | 512+128 | 2,111 ms | llamafile 0.9.2 |
| Q4_K - Medium | 40.6 | 579 | LocalScore ↗ | 512+128 | 2,244 ms | llamafile 0.9.2 |
| Q4_K - Medium | 39.7 | 606 | LocalScore ↗ | 512+128 | 2,187 ms | llamafile 0.9.2 |
| Q4_K - Medium | 39.1 | 596 | LocalScore ↗ | 512+128 | 2,209 ms | llamafile 0.9.2 |
| Q4_K - Medium | 39.0 | 578 | LocalScore ↗ | 512+128 | 2,315 ms | llamafile 0.9.2 |
| Q4_K - Medium | 38.7 | 579 | LocalScore ↗ | 4,352 | 2,239 ms | llamafile 0.9.2 |
| Q4_K - Medium | 22.1 | 369 | LocalScore ↗ | 512+128 | 3,895 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 (8-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.