← Back to community benchmarks

Qwen3.6-35B-A3B

M5 Pro (20c) · 48 GB · 4bit · 2026-04-20
Performance
32k
tokens
1,804
PP tok/s
43.2
TG tok/s
18165
TTFT (ms)
22.4
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 48 GB
GPU Cores 20
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 32,768
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,503 95.8 19.2 GB view
4k 2,179 93.7 20.0 GB view
8k 2,253 84.9 20.4 GB view
16k 2,134 69.9 21.0 GB view
32k 1,804 43.2 22.4 GB current
64k 1,275 27.5 25.0 GB view