← Back to community benchmarks

Qwen3.6-35B-A3B

M3 Ultra (60c) · 96 GB · 4bit · 2026-04-23

Performance

64k

tokens

1,286

PP tok/s

45.9

TG tok/s

50981

TTFT (ms)

25.0

Peak mem (GB)

Hardware

Chip M3 Ultra (60c)

Memory 96 GB

GPU Cores 60

Software

oMLX v0.3.8.dev1

macOS macOS 26.4.1

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
4k	2,047	97.9	20.0 GB	view
64k	1,286	45.9	25.0 GB	current
128k	837.6	27.3	30.4 GB	view