← Back to community benchmarks

Qwen3-Coder-30B-A3B-Instruct-8bit-DWQ-lr2e7

M5 Max (40c) · 128 GB · 8bit · 2026-03-17

Performance

4k

tokens

2,473

PP tok/s

75.9

TG tok/s

1657

TTFT (ms)

33.1

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.2.14

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	2,053	89.0	32.6 GB	view
4k	2,473	75.9	33.1 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	89.0	1.00×
2×	126.9	1.43×
4×	172.3	1.94×