← Back to community benchmarks

Qwen2.5-VL-32B-Instruct

M3 Ultra (60c) · 96 GB · 8bit · 2026-05-11

Performance

4k

tokens

261.9

PP tok/s

17.6

TG tok/s

15637

TTFT (ms)

34.6

Peak mem (GB)

Hardware

Chip M3 Ultra (60c)

Memory 96 GB

GPU Cores 60

Software

oMLX v0.3.8

macOS macOS 26.4.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	255.1	18.3	33.1 GB	view
4k	261.9	17.6	34.6 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	18.3	1.00×
2×	29.3	1.60×
4×	37.0	2.02×