Skip to content

Aria-UI

Ground GUI instructions with a vision-only model

Visit official site

What Aria-UI is

The model is a large multimodal model built for visual grounding of GUI instructions.

What you can do with it

  • Understand grounding instructions in varied formats
  • Use history for context-aware grounding
  • Adapt to different planning agents

Pros & cons

✓ What we like

  • Multiple instruction formats are handled
  • Context comes from history input

! What to watch out for

  • The approach relies on visual input only
  • Local use needs datasets and checkpoints

FAQ

What inputs are used?

A purely visual approach.

How large is it?

3.9B active parameters per token.

Last reviewed: 2026-11-15

More AI design tools tools

View all →

How we review