← Back to community benchmarks

Qwen3.5-0.8B

M4 Pro (16c) · 24 GB · 4bit · 2026-03-10

Performance

4k

tokens

3,700

PP tok/s

278.9

TG tok/s

1107

TTFT (ms)

1.9

Peak mem (GB)

Hardware

Chip M4 Pro (16c)

Memory 24 GB

GPU Cores 16

Software

oMLX v0.2.7

macOS macOS 26.3

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	3,705	303.8	1.7 GB	view
4k	3,700	278.9	1.9 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	303.8	1.00×
2×	488.2	1.61×
4×	610.9	2.01×