← Back to community benchmarks

Qwen3.6-35B-A3B-oQ8

M1 Max (32c) · 64 GB · 8bit · 2026-04-17

Performance

8k

tokens

991.5

PP tok/s

50.6

TG tok/s

8262

TTFT (ms)

36.5

Peak mem (GB)

Hardware

Chip M1 Max (32c)

Memory 64 GB

GPU Cores 32

Software

oMLX v0.3.6

macOS macOS 26.4.1

Context 8,192

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	696.2	54.9	35.4 GB	view
4k	978.9	52.5	36.2 GB	view
8k	991.5	50.6	36.5 GB	current
16k	938.5	46.5	37.1 GB	view
32k	816.6	39.8	38.5 GB	view
64k	584.5	29.6	41.2 GB	view