← Back to community benchmarks

Qwen3.6-27B-oQ4-mtp

M3 Ultra (60c) · 96 GB · 4bit · 2026-05-11
Performance
1k
tokens
302.2
PP tok/s
42.6
TG tok/s
3389
TTFT (ms)
16.8
Peak mem (GB)
Hardware
Chip M3 Ultra (60c)
Memory 96 GB
GPU Cores 60
Software
oMLX v0.3.9.dev1
macOS macOS 26.4.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 302.2 42.6 16.8 GB current
4k 318.4 43.2 18.2 GB view
8k 317.1 41.5 19.2 GB view
16k 308.7 38.8 20.7 GB view
Batching Results
Batch Size TG tok/s Speedup
42.6 1.00×
45.7 1.07×
62.5 1.47×