← Back to community benchmarks

Qwen3.5-4B

M4 (10c) · 16 GB · 4bit · 2026-04-18

Performance

1k

tokens

291.7

PP tok/s

105.8

TG tok/s

3511

TTFT (ms)

4.3

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 16 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	291.7	105.8	4.3 GB	current
4k	309.1	35.1	3.9 GB	view
8k	337.4	32.6	4.3 GB	view