← Back to community benchmarks

tfjack-Qwen3.6-35B-A3B-oQ8-fp16-mtp

M3 Ultra (60c) · 96 GB · 8bit · 2026-05-18

Performance

1k

tokens

1,547

PP tok/s

93.2

TG tok/s

662

TTFT (ms)

36.2

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 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,547	93.2	36.2 GB	current
64k	1,372	63.0	42.0 GB	view
128k	861.7	47.7	48.7 GB	view
195k	614.5	38.7	55.6 GB	view