← Back to community benchmarks

Qwen3.6-35B-A3B-Hereticmixed-6.4bit

M5 Pro (20c) · 48 GB · 6bit · 2026-05-01

Performance

1k

tokens

1,416

PP tok/s

70.4

TG tok/s

723

TTFT (ms)

26.6

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 48 GB

GPU Cores 20

Software

oMLX v0.3.8

macOS macOS 26.4.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,416	70.4	26.6 GB	current
4k	2,106	68.5	27.4 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	70.4	1.00×
2×	102.9	1.46×
4×	143.2	2.03×