← Back to community benchmarks

Qwen3.6-35B-A3B

M5 Pro (20c) · 64 GB · 4bit · 2026-07-04
Performance
195k
tokens
579.8
PP tok/s
34.5
TG tok/s
344925
TTFT (ms)
36.1
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.4.4
macOS macOS 26.5.1
Context 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,269 104.8 19.3 GB view
4k 2,029 96.6 20.0 GB view
8k 2,085 97.1 20.4 GB view
16k 2,044 89.8 21.0 GB view
32k 1,723 78.4 22.4 GB view
64k 1,258 61.4 25.0 GB view
128k 808.0 44.4 30.4 GB view
195k 579.8 34.5 36.1 GB current
Batching Results
Batch Size TG tok/s Speedup
104.8 1.00×
146.3 1.40×
190.5 1.82×
233.2 2.23×