← Back to community benchmarks

Qwen3.5-9Blm

M4 (10c) · 24 GB · 4bit · 2026-05-10

Performance

4k

tokens

186.4

PP tok/s

20.7

TG tok/s

21972

TTFT (ms)

6.5

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 24 GB

GPU Cores 10

Software

oMLX v0.3.8

macOS macOS 26.4.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	172.2	75.9	7.0 GB	view
4k	186.4	20.7	6.5 GB	current
8k	195.8	20.2	6.9 GB	view
16k	567.7	18.9	7.6 GB	view
32k	519.5	20.0	8.3 GB	view
64k	454.8	17.2	9.4 GB	view