← Back to community benchmarks

Qwen3.6-35B-A3B-oQ4-mtp

M2 Pro (16c) · 32 GB · 4bit · 2026-05-18
Performance
1k
tokens
291.2
PP tok/s
44.0
TG tok/s
3517
TTFT (ms)
20.4
Peak mem (GB)
Hardware
Chip M2 Pro (16c)
Memory 32 GB
GPU Cores 16
Software
oMLX v0.3.9.dev2
macOS macOS 26.4.1
Context 1,024
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 291.2 44.0 20.4 GB current
4k 335.7 44.1 21.2 GB view
8k 333.2 41.2 21.7 GB view
16k 316.3 38.4 22.4 GB view
32k 282.3 29.6 23.9 GB view
Batching Results
Batch Size TG tok/s Speedup
44.0 1.00×
65.8 1.50×