← Back to community benchmarks

Qwen3.5-4B

M4 (10c) · 16 GB · 4bit · 2026-04-28

Performance

4k

tokens

393.3

PP tok/s

35.5

TG tok/s

10413

TTFT (ms)

3.9

Peak mem (GB)

Hardware

Chip M4 (10c)

Memory 16 GB

GPU Cores 10

Software

oMLX v0.3.8.dev3

macOS macOS 26.4.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	379.0	37.4	3.3 GB	view
4k	393.3	35.5	3.9 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	37.4	1.00×
2×	69.6	1.86×
4×	90.2	2.41×
8×	89.4	2.39×