← Back to community benchmarks

Qwen3.5-4B-OptiQ

M5 Max (40c) · 128 GB · 4bit · 2026-06-05

Performance

8k

tokens

3,219

PP tok/s

82.4

TG tok/s

2545

TTFT (ms)

5.2

Peak mem (GB)

Hardware

Chip M5 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.4.1

macOS macOS 26.5.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	2,489	89.6	4.2 GB	view
4k	2,544	75.2	4.8 GB	view
8k	3,219	82.4	5.2 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	89.6	1.00×
2×	165.4	1.85×
4×	91.4	1.02×