← Back to community benchmarks

Qwen3.6-35B-A3B-Uncensored-HauhauCS-Aggressive-OptiQ-6bpw

M5 Pro (20c) · 64 GB · 4bit · 2026-07-21
Performance
195k
tokens
520.2
PP tok/s
30.3
TG tok/s
384484
TTFT (ms)
39.4
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 200,000
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 2,057 82.8 19.7 GB view
4k 2,460 80.9 20.5 GB view
8k 2,384 78.0 21.0 GB view
16k 2,183 73.2 21.7 GB view
32k 1,821 65.8 23.6 GB view
64k 1,296 53.9 26.2 GB view
128k 775.7 37.7 32.6 GB view
195k 520.2 30.3 39.4 GB current
Batching Results
Batch Size TG tok/s Speedup
82.8 1.00×
135.9 1.64×
198.1 2.39×
373.1 4.51×