← Back to community benchmarks

Qwen3.5-35B-A3B-5bit

M2 Pro (19c) · 32 GB · 5bit · 2026-03-11

Performance

1k

tokens

351.6

PP tok/s

48.9

TG tok/s

2912

TTFT (ms)

24.1

Peak mem (GB)

Hardware

Chip M2 Pro (19c)

Memory 32 GB

GPU Cores 19

Software

oMLX v0.2.7

macOS macOS 26.3.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	351.6	48.9	24.1 GB	current
4k	364.1	46.5	24.2 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	48.9	1.00×
2×	57.8	1.18×
4×	66.0	1.35×