← Back to community benchmarks

Qwen3-VL-30B-A3B-Instruct

M3 Ultra (80c) · 512 GB · 4bit · 2026-03-12

Performance

1k

tokens

1,909

PP tok/s

71.1

TG tok/s

536

TTFT (ms)

17.6

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 512 GB

GPU Cores 80

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	1,909	71.1	17.6 GB	current
4k	2,335	68.0	18.1 GB	view

Batching Results

Batch Size	TG tok/s	Speedup
1×	71.1	1.00×
2×	118.0	1.66×
4×	197.7	2.78×