← Back to community benchmarks

Qwen3.6-35B-A3B-oQ6-fp16-mtp

M3 Ultra (60c) · 96 GB · 6bit · 2026-05-18

Performance

128k

tokens

871.5

PP tok/s

49.0

TG tok/s

150391

TTFT (ms)

40.8

Peak mem (GB)

Hardware

Chip M3 Ultra (60c)

Memory 96 GB

GPU Cores 60

Software

oMLX v0.3.9.dev2

macOS macOS 26.4

Context 131,072

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,539	94.5	28.3 GB	view
64k	1,364	63.6	34.1 GB	view
128k	871.5	49.0	40.8 GB	current
195k	612.7	38.8	47.7 GB	view