← Back to community benchmarks

Qwen3.6-35B-A3B-4bit-MTPLX-Optimized-Speed

M3 Pro (14c) · 36 GB · 4bit · 2026-05-10

Performance

16k

tokens

144.8

PP tok/s

32.1

TG tok/s

113175

TTFT (ms)

24.9

Peak mem (GB)

Hardware

Chip M3 Pro (14c)

Memory 36 GB

GPU Cores 14

Software

oMLX v0.3.9.dev1

macOS macOS 26.4.1

Context 16,384

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	174.2	36.0	21.9 GB	view
4k	173.9	35.1	24.5 GB	view
8k	155.9	33.2	24.7 GB	view
16k	144.8	32.1	24.9 GB	current
32k	125.1	27.6	25.4 GB	view