M4 Pro (16-core GPU) — benchmark record
M4 Pro (16-core GPU) local LLM benchmarks across 24 GB, 48 GB, 64 GB RAM tiers on Apple Silicon Mac. 8 published rows across 3 models with explicit evidence state and RAM-tier comparison. Peak published speed is 111.9 tok/s.
8Benchmark rows
3Models tested
3RAM configurations
111.9Fastest avg tok/s
RAM configurations
Each configuration differs only in unified memory. More RAM = larger models fit. Throughput is similar across RAM tiers at the same model size.
All benchmark rows — M4 Pro (16-core GPU)
Sorted by avg tok/s descending. Click source badge to see original measurement.
| Chip (RAM) | Model | Quant | Avg tok/s | Runtime | Source |
|---|---|---|---|---|---|
| M4 Pro (16-core GPU, 64 GB) | Llama 3.2 1B | Q4_K - Medium | 111.9 tok/s | llamafile | ref |
| M4 Pro (16-core GPU, 48 GB) | Llama 3.2 1B | Q4_K - Medium | 111.0 tok/s | llamafile | ref |
| M4 Pro (16-core GPU, 24 GB) | Llama 3.2 1B | Q4_K - Medium | 110.9 tok/s | llamafile | ref |
| M4 Pro (16-core GPU, 24 GB) | Llama 3.1 8B | Q4_K - Medium | 30.5 tok/s | llamafile | ref |
| M4 Pro (16-core GPU, 48 GB) | Llama 3.1 8B | Q4_K - Medium | 30.2 tok/s | llamafile | ref |
| M4 Pro (16-core GPU, 48 GB) | Qwen 2.5 14B | Q4_K - Medium | 16.8 tok/s | llamafile | ref |
| M4 Pro (16-core GPU, 64 GB) | Qwen 2.5 14B | Q4_K - Medium | 16.1 tok/s | llamafile | ref |
| M4 Pro (16-core GPU, 24 GB) | Qwen 2.5 14B | Q4_K - Medium | 15.2 tok/s | llamafile | ref |
Models tested on M4 Pro (16-core GPU)
Data
benchmarks.json — full dataset · chips.json — chip summaries · benchmarks.csv — CSV export
Data from in-house lab measurements plus community-published benchmarks. See all chip families →