← Back to community benchmarks

Qwen3.5-9B

M3 Max · 16 GB · 4bit · 2026-03-23

Performance

1k

tokens

505.4

PP tok/s

53.6

TG tok/s

2026

TTFT (ms)

6.5

Peak mem (GB)

Hardware

Chip M3 Max

Memory 16 GB

Software

oMLX v0.2.19

macOS macOS 26.3.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	505.4	53.6	6.5 GB	current
4k	538.9	49.9	7.3 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	53.6	1.00×
2×	92.9	1.73×
4×	113.2	2.11×