← Back to community benchmarks

unsloth--Qwen3.6-35B-A3B-UD

M5 Pro (20c) · 64 GB · 4bit · 2026-07-18
Performance
16k
tokens
1,415
PP tok/s
65.8
TG tok/s
11576
TTFT (ms)
21.7
Peak mem (GB)
Hardware
Chip M5 Pro (20c)
Memory 64 GB
GPU Cores 20
Software
oMLX v0.5.1
macOS macOS 26.5.2
Context 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,963 73.7 20.4 GB view
4k 2,441 70.7 21.2 GB view
8k 2,363 69.3 21.6 GB view
16k 1,415 65.8 21.7 GB current
32k 674.5 31.1 28.1 GB view
64k 623.7 50.4 22.9 GB view