← Back to community benchmarks

Qwen3-Coder-30B-A3B-Instruct

M3 Max (40c) · 128 GB · 4bit · 2026-03-14
Performance
4k
tokens
1,204
PP tok/s
71.8
TG tok/s
3404
TTFT (ms)
17.1
Peak mem (GB)
Hardware
Chip M3 Max (40c)
Memory 128 GB
GPU Cores 40
Software
oMLX v0.2.8
macOS macOS 26.3.1
Context 4,096
Performance by Context Length
Context PP tok/s TG tok/s Peak Mem
1k 1,175 80.5 16.6 GB view
4k 1,204 71.8 17.1 GB current
Batching Results
Batch Size TG tok/s Speedup
1× 80.5 1.00×
2× 133.6 1.66×
4× 155.8 1.94×
8× 210.7 2.62×