← Back to community benchmarks

Qwen3.6-35B-A3B-OptiQ

M3 Ultra (80c) · 96 GB · 4bit · 2026-07-12

Performance

64k

tokens

1,609

PP tok/s

68.2

TG tok/s

40742

TTFT (ms)

24.1

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 96 GB

GPU Cores 80

Software

oMLX v0.5.0

macOS macOS 26.5.2

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
32k	2,112	77.2	23.0 GB	view
64k	1,609	68.2	24.1 GB	current
128k	1,087	57.3	25.3 GB	view