← Back to community benchmarks

Qwen3.6-35B-A3BoQ4

M3 Max (30c) · 36 GB · 4bit · 2026-04-17
Performance
4k
tokens
1,200
PP tok/s
71.5
TG tok/s
3413
TTFT (ms)
20.8
Peak mem (GB)
Hardware
Chip M3 Max (30c)
Memory 36 GB
GPU Cores 30
Software
oMLX v0.3.6
macOS macOS 26.4.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 975.7 74.8 20.0 GB view
4k 1,200 71.5 20.8 GB current
Batching Results
Batch Size TG tok/s Speedup
74.8 1.00×
116.1 1.55×
152.1 2.03×