← Back to community benchmarks

Qwen3-Coder-Next

M5 Pro (20c) · 64 GB · 4bit · 2026-03-22

Performance

4k

tokens

1,653

PP tok/s

66.3

TG tok/s

2478

TTFT (ms)

44.0

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 64 GB

GPU Cores 20

Software

oMLX v0.2.20.dev1

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,221	69.7	43.1 GB	view
4k	1,653	66.3	44.0 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	69.7	1.00×
2×	110.5	1.59×
4×	157.3	2.26×