M3 Ultra (60-core GPU, 96 GB) — LLM Benchmarks

Measured LLM inference benchmarks for M3 Ultra (60-core GPU, 96 GB). Tokens per second across 1 model and multiple quantizations. Real runs, not estimates.

1 Benchmark rows

1 Models tested

34.4 Fastest avg tok/s (Qwen 2.5 14B Instruct)

0 Factory-lab verified rows

Benchmark results for M3 Ultra (60-core GPU, 96 GB)

Rows sorted by avg tok/s descending. Click source badge to see original measurement page.

Model	Quant	RAM req.	Context	Avg tok/s	Prompt tok/s	Runtime	Source
Qwen 2.5 14B Instruct	Q4_K - Medium	—	—	34.4 tok/s	444.6 tok/s	—	ref

Models tested on this chip

Qwen 2.5 14B Instruct

Data

benchmarks.json — full dataset · chips.json — chip summaries · benchmarks.csv — CSV export

Want to contribute a benchmark for this chip? Data sourced from factory lab measurements and community reference runs. See all chips →