← Back to community benchmarks

Qwen3.5-9Blm

M5 Pro (20c) · 64 GB · 4bit · 2026-04-12

Performance

4k

tokens

1,527

PP tok/s

57.6

TG tok/s

2682

TTFT (ms)

6.3

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 64 GB

GPU Cores 20

Software

oMLX v0.3.5.dev1

macOS macOS 26.3.2

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,260	59.2	5.6 GB	view
4k	1,527	57.6	6.3 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	59.2	1.00×
2×	103.3	1.74×
4×	131.6	2.22×