M5 Pro — benchmark record

M5 Pro local LLM benchmarks across 24 GB, 64 GB RAM tiers on Apple Silicon Mac. 6 published rows across 6 models with explicit evidence state and RAM-tier comparison. Peak published speed is 92.0 tok/s.

6Benchmark rows

6Models tested

2RAM configurations

92.0Fastest 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.

24 GBM5 Pro (24 GB)2 rows · 2 models92.0 tok/s peak 64 GBM5 Pro (64 GB)4 rows · 4 models41.9 tok/s peak

All benchmark rows — M5 Pro

Sorted by avg tok/s descending. Click source badge to see original measurement.

Chip (RAM)	Model	Quant	RAM req.	Avg tok/s	Prompt tok/s	Runtime	Source
M5 Pro (24 GB)	Gemma 4 E4B	Q4_K - Medium	—	92.0 tok/s	—	Ollama	ref
M5 Pro (64 GB)	Qwen3.5-35B-A3B	Q4_K_L	27.2 GB	41.9 tok/s	—	llama.cpp	ref
M5 Pro (24 GB)	Gemma 4 26B-A4B	Q4_K - Medium	—	35.0 tok/s	—	Ollama	ref
M5 Pro (64 GB)	Qwen3.5-9B	Q4_K - Medium	13.8 GB	25.0 tok/s	—	llama.cpp	ref
M5 Pro (64 GB)	Qwen3.5-122B-A10B	IQ1_M	40.8 GB	18.0 tok/s	—	llama.cpp	ref
M5 Pro (64 GB)	Qwen3.5-27B	Q4_K - Medium	24.9 GB	10.0 tok/s	—	llama.cpp	ref

Models tested on M5 Pro

Gemma 4 26B-A4B Gemma 4 E4B Qwen3.5-9B Qwen3.5-27B Qwen3.5-35B-A3B Qwen3.5-122B-A10B

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 →