← Back to community benchmarks

Qwen3.6-35B-A3BVLM-oQ8

M1 Ultra (64c) · 64 GB · 8bit · 2026-04-24

Performance

16k

tokens

1,758

PP tok/s

52.5

TG tok/s

9322

TTFT (ms)

37.1

Peak mem (GB)

Hardware

Chip M1 Ultra (64c)

Memory 64 GB

GPU Cores 64

Software

oMLX v0.3.6

macOS macOS 26.4.1

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,084	193.4	36.7 GB	view
4k	302.9	61.4	36.2 GB	view
8k	1,807	59.3	36.5 GB	view
16k	1,758	52.5	37.1 GB	current
32k	1,478	38.2	38.5 GB	view
64k	926.4	25.4	41.2 GB	view