← Back to community benchmarks

Qwen3.5-9B

M3 Pro (18c) · 18 GB · 4bit · 2026-03-14

Performance

1k

tokens

335.7

PP tok/s

26.6

TG tok/s

3051

TTFT (ms)

6.7

Peak mem (GB)

Hardware

Chip M3 Pro (18c)

Memory 18 GB

GPU Cores 18

Software

oMLX v0.2.11

macOS macOS 26.3.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	335.7	26.6	6.7 GB	current
4k	338.9	25.9	7.0 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	26.6	1.00×
2×	53.1	2.00×
4×	79.9	3.00×