← Back to community benchmarks

Qwen3.6-35B-A3B

M5 Pro (20c) · 48 GB · 4bit · 2026-06-05

Performance

4k

tokens

1,953

PP tok/s

26.1

TG tok/s

2097

TTFT (ms)

24.0

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 48 GB

GPU Cores 20

Software

oMLX v0.4.1

macOS macOS 26.5.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,463	96.9	19.3 GB	view
4k	1,953	26.1	24.0 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	96.9	1.00×
2×	76.4	0.79×
4×	137.6	1.42×