← Back to community benchmarks

Qwen3-Coder-30B-A3B-Instruct

M1 Ultra (48c) · 64 GB · 4bit · 2026-03-13

Performance

1k

tokens

658.9

PP tok/s

58.6

TG tok/s

1554

TTFT (ms)

16.6

Peak mem (GB)

Hardware

Chip M1 Ultra (48c)

Memory 64 GB

GPU Cores 48

Software

oMLX v0.2.8

macOS macOS 26.2

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	658.9	58.6	16.6 GB	current
4k	681.1	49.3	17.1 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	58.6	1.00×
2×	91.6	1.56×
4×	121.7	2.08×