← Back to community benchmarks

Qwen3.5-4B

M3 (10c) · 24 GB · 4bit · 2026-03-11

Performance

1k

tokens

181.6

PP tok/s

12.8

TG tok/s

5639

TTFT (ms)

3.9

Peak mem (GB)

Hardware

Chip M3 (10c)

Memory 24 GB

GPU Cores 10

Software

oMLX v0.2.7

macOS macOS 26.2

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	181.6	12.8	3.9 GB	current
4k	183.6	13.0	4.2 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	12.8	1.00×
2×	23.0	1.80×
4×	26.1	2.04×