← Back to community benchmarks

Qwen3-VL-30B-A3B-Instruct

M3 Ultra (80c) · 256 GB · unknown · 2026-03-10

Performance

4k

tokens

2,141

PP tok/s

28.0

TG tok/s

1913

TTFT (ms)

59.0

Peak mem (GB)

Hardware

Chip M3 Ultra (80c)

Memory 256 GB

GPU Cores 80

Software

oMLX v0.2.6

macOS macOS 26.3.1

Context 4,096

Performance by Context Length

Context	PP tok/s	TG tok/s	Peak Mem
1k	903.1	34.7	58.4 GB	view
4k	2,141	28.0	59.0 GB	current

Batching Results

Batch Size	TG tok/s	Speedup
1×	34.7	1.00×
2×	42.6	1.23×
4×	73.2	2.11×