← Back to community benchmarks

Qwen3.5-9B-OptiQ

M5 (10c) · 32 GB · 4bit · 2026-04-25

Performance

1k

tokens

699.5

PP tok/s

25.1

TG tok/s

1464

TTFT (ms)

6.7

Peak mem (GB)

Hardware

Chip M5 (10c)

Memory 32 GB

GPU Cores 10

Software

oMLX v0.3.6

macOS macOS 26.3.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	699.5	25.1	6.7 GB	current
4k	776.0	24.4	7.3 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	25.1	1.00×
2×	50.8	2.02×
4×	70.0	2.79×