← Back to community benchmarks

Qwen3-VL-8B-Instruct

M4 Pro (20c) · 64 GB · 8bit · 2026-03-11

Performance

1k

tokens

442.6

PP tok/s

28.8

TG tok/s

2314

TTFT (ms)

9.9

Peak mem (GB)

Hardware

Chip M4 Pro (20c)

Memory 64 GB

GPU Cores 20

Software

oMLX v0.2.8b1

macOS macOS 26.2

Context 1,024

Batching Results

Batch Size	TG tok/s	Speedup
1×	28.8	1.00×
2×	54.0	1.88×
4×	67.5	2.34×