← Back to community benchmarks

Qwen3.5-9B

M4 (10c) · 16 GB · 4bit · 2026-03-10

Performance

16k

tokens

204.2

PP tok/s

16.3

TG tok/s

80230

TTFT (ms)

7.5

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 16 GB

GPU Cores 10

Software

oMLX v0.2.7

macOS macOS 26.2

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
16k	204.2	16.3	7.5 GB	current
32k	187.5	12.4	8.5 GB	view