← Back to community benchmarks

Qwen3.5-4B

M3 Ultra (60c) · 96 GB · unknown · 2026-03-10

Performance

4k

tokens

1,539

PP tok/s

17.0

TG tok/s

2661

TTFT (ms)

12.1

Peak mem (GB)

Hardware

Chip M3 Ultra (60c)

Memory 96 GB

GPU Cores 60

Software

oMLX v0.2.6

macOS macOS 15.3

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,250	17.3	12.0 GB	view
4k	1,539	17.0	12.1 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	17.3	1.00×
2×	21.7	1.25×
4×	42.8	2.47×