← Back to community benchmarks

qwen3.6-35b-a3b-unsloth

M3 Max (40c) · 128 GB · 4bit · 2026-04-27
Performance
4k
tokens
1,446
PP tok/s
78.3
TG tok/s
2833
TTFT (ms)
21.2
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.3.5.dev2
macOS macOS 15.7.2
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,159 80.6 20.4 GB view
4k 1,446 78.3 21.2 GB current
8k 1,349 71.8 21.5 GB view
Batching Results
Batch Size TG tok/s Speedup
80.6 1.00×
132.3 1.64×
191.0 2.37×