← Back to community benchmarks

Qwen3.5-2B

M4 Max (40c) · 128 GB · 6bit · 2026-03-23

Performance

4k

tokens

3,746

PP tok/s

179.4

TG tok/s

1093

TTFT (ms)

3.5

Peak mem (GB)

Hardware

Chip M4 Max (40c)

Memory 128 GB

GPU Cores 40

Software

oMLX v0.2.20rc1

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	3,492	202.6	3.1 GB	view
4k	3,746	179.4	3.5 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	202.6	1.00×
2×	335.5	1.66×
4×	410.5	2.03×