M2 Ultra (GPU count not published) — benchmark record

M2 Ultra (GPU count not published) local LLM benchmarks across 128 GB RAM tiers on Apple Silicon Mac. 3 published rows across 2 models with explicit evidence state and RAM-tier comparison. Peak published speed is 120.4 tok/s.

3Benchmark rows

2Models tested

1RAM configurations

120.4Fastest 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.

128 GBM2 Ultra (GPU count not published, 128 GB)3 rows · 2 models120.4 tok/s peak

All benchmark rows — M2 Ultra (GPU count not published)

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
M2 Ultra (GPU count not published, 128 GB)	Llama 3.2 1B	Q4_K - Medium	—	120.4 tok/s	659.8 tok/s	llamafile	ref
M2 Ultra (GPU count not published, 128 GB)	Qwen3.5-27B	8bit	—	27.1 tok/s	—	MLX	ref
M2 Ultra (GPU count not published, 128 GB)	Qwen3.5-27B	8bit	—	20.6 tok/s	—	MLX	ref

Models tested on M2 Ultra (GPU count not published)

Llama 3.2 1B Qwen3.5-27B

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 →