← Back to community benchmarks

Qwen3.6-35B-A3B

M5 Pro (20c) · 64 GB · 4bit · 2026-07-13

Performance

64k

tokens

571.2

PP tok/s

38.2

TG tok/s

114740

TTFT (ms)

21.2

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 64 GB

GPU Cores 20

Software

oMLX v0.5.1

macOS macOS 26.5.2

Context 65,536

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,576	82.2	19.3 GB	view
4k	1,837	83.8	20.0 GB	view
64k	571.2	38.2	21.2 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	82.2	1.00×
2×	143.8	1.75×
4×	162.0	1.97×