M3 Ultra vs M4 Max
Signed, community-submitted decode tok/s, prefill, and TTFT for M3 Ultra vs M4 Max — every number links to the run it came from.
Verdict
M3 Ultra and M4 Max both have submitted runs, but no single model has been measured on both yet — so there is no apples-to-apples row. Each side's measured decode tok/s is listed separately below (peaks of 192 and 113 tok/s). Submit a shared workload to turn this into a direct comparison.
M3 Ultra benchmarks — no shared model with M4 Max yet
M4 Max benchmarks — no shared model with M3 Ultra yet
| Model | decode tok/s | Workload | Run |
|---|---|---|---|
| qwen3-coder-bench-32k | 113.3tok/s | chat-short | r_roktphpc--8 |
| qwen3-coder | 109.9tok/s | concurrent-decode | r_r0di2hkku1h |
See also: M3 Ultra benchmarks · M4 Max benchmarks · All hardware · All models · Methodology