← Back to community benchmarks

Qwen3.6-35B-A3B

M5 Pro (20c) · 48 GB · 4bit · 2026-04-26
Performance
16k
tokens
2,116
PP tok/s
79.3
TG tok/s
7743
TTFT (ms)
26.0
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 16,384
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,487 97.0 24.3 GB view
4k 2,165 91.3 25.1 GB view
8k 2,238 87.0 25.4 GB view
16k 2,116 79.3 26.0 GB current
Batching Results
Batch Size TG tok/s Speedup
97.0 1.00×
192.4 1.98×
379.6 3.91×