← Back to community benchmarks

Qwen3.5-35B-A3B

M5 Pro (20c) · 64 GB · 4bit · 2026-03-18

Performance

1k

tokens

1,968

PP tok/s

92.0

TG tok/s

520

TTFT (ms)

20.1

Peak mem (GB)

Hardware

Chip M5 Pro (20c)

Memory 64 GB

GPU Cores 20

Software

oMLX v0.2.18

macOS macOS 26.3.1

Context 1,024

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	1,968	92.0	20.1 GB	current
4k	2,328	88.3	21.0 GB	view
8k	2,266	83.1	21.4 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	92.0	1.00×
2×	147.8	1.61×
4×	196.6	2.14×