← Back to community benchmarks

Qwen3.6-27B-oQ4-fp16-mtp

M3 Max (30c) · 36 GB · 4bit · 2026-05-20
Performance
1k
tokens
166.5
PP tok/s
25.1
TG tok/s
6150
TTFT (ms)
16.7
Peak mem (GB)
Hardware
Chip M3 Max (30c)
Memory 36 GB
GPU Cores 30
Software
oMLX v0.3.9rc1
macOS macOS 26.4.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 166.5 25.1 16.7 GB current
32k 85.0 9.8 22.5 GB view
Batching Results
Batch Size TG tok/s Speedup
25.1 1.00×
15.8 0.63×
19.7 0.78×