Apple M4 (8-core GPU) × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on Apple M4 (8-core GPU)
66.3best gen tok/s
904best prompt tok/s
4,352max context tested
6results
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 | 66.3 | 904 | LocalScore ↗ | 512+128 | 1,430 ms | llamafile 0.9.2 |
| Q4_K - Medium | 66.3 | 902 | LocalScore ↗ | 4,352 | 1,437 ms | llamafile 0.9.2 |
| Q4_K - Medium | 66.2 | 895 | LocalScore ↗ | 512+128 | 1,442 ms | llamafile 0.9.2 |
| Q4_K - Medium | 65.8 | 903 | LocalScore ↗ | 512+128 | 1,440 ms | llamafile 0.9.2 |
| Q4_K - Medium | 65.6 | 898 | LocalScore ↗ | 512+128 | 1,443 ms | llamafile 0.9.2 |
| Q4_K - Medium | 65.5 | 886 | LocalScore ↗ | 512+128 | 1,451 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 (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.