← Back to community benchmarks

Qwen3-Coder-30B-A3B-Instruct

M3 Ultra (80c) · 512 GB · 8bit · 2026-04-18

Performance

8k

tokens

1,996

PP tok/s

56.8

TG tok/s

4103

TTFT (ms)

31.7

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

Software

oMLX v0.3.6

macOS macOS 26.5

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,805	277.1	31.3 GB	view
4k	939.1	51.1	31.3 GB	view
8k	1,996	56.8	31.7 GB	current
16k	1,613	43.4	32.4 GB	view