Documented entries use openly licensed vendor imagery, credited · Concept entries are illustrative
AgentUI Gallery

About

A small reference for how AI agents ask, plan, and act.

AgentUI Gallery is an independent editorial collection of real AI agent interfaces. Each entry is one screen from a shipped product, paired with one design insight you can apply to your own work. It’s deliberately small: we would rather publish a few entries that hold up in a design review than hundreds of thin ones.

It is curated by Summer Ordonez, a product and interaction designer. It isn’t affiliated with, sponsored by, or endorsed by any company whose product appears here.

Documented vs. Our read

Every entry separates two kinds of statement:

  • Documented: facts taken from the vendor’s own documentation or release notes, each shown with a quote and a link to where it appears, plus a description of what is visible in the image.
  • Our read: our design interpretation: why the pattern works, its trade-off, and what to borrow. Opinions are labeled as opinions.

Editorial standards

  • Every documented entry uses an image we are allowed to publish (an open license or written permission), credited with its license.
  • Images are cropped or resized only, never edited. Each one is labeled Unmodified or Cropped.
  • Every factual claim is re-checked against its source before publishing; history items carry a date and a source URL.
  • Each insight is specific, names a trade-off, and is something a designer could act on.
  • Concept entries and timelines are illustrative teaching examples, labeled as such, and never presented as product history.

Corrections and removal requests: see Image credits & takedown.