← Back to community benchmarks

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

M1 Max (32c) · 32 GB · 4bit · 2026-05-20
Performance
8k
tokens
145.6
PP tok/s
21.6
TG tok/s
56272
TTFT (ms)
19.2
Peak mem (GB)
Hardware
Chip M1 Max (32c)
Memory 32 GB
GPU Cores 32
Software
oMLX v0.3.9rc1
macOS macOS 26.5
Context 8,192
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 123.4 23.3 16.7 GB view
4k 147.9 22.3 18.2 GB view
8k 145.6 21.6 19.2 GB current
Batching Results
Batch Size TG tok/s Speedup
1× 23.3 1.00×
2× 41.6 1.79×