← Back to community benchmarks

Qwen3.5-9B-OptiQ

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

Performance

1k

tokens

1,233

PP tok/s

51.8

TG tok/s

831

TTFT (ms)

6.8

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 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,233	51.8	6.8 GB	current
4k	1,519	50.3	7.6 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	51.8	1.00×
2×	94.0	1.81×
4×	129.2	2.49×