Hollow Studio is the full-screen home for serious work with the AI: a conversation panel on the left, a live preview and artifact inspector on the right. Long-form projects, software development, design generation, cross-app automated workflows: Hollow Studio carries all of it.


Review a document from a conversation.
Group conversations into Spaces, rename and pin them, and find an old chat from the topbar search. Fork a conversation from any assistant reply to try a different direction without touching the original. Pull in files from the Space, a URL or an earlier conversation with an @ mention, and set standing instructions for a whole Space.
Documents and files the assistant makes land as versioned artifacts, such as source code, Markdown, an HTML page, a spreadsheet, a Word document, a diagram or SVG graphics. Each artifact keeps its versions, so you can compare revisions and restore an earlier one.
For a renderable artifact, the Preview panel puts a live browser beside your conversation and renders it as the code is written. Static web pages and React applications bundle client-side and become real, interactive components you can click through, not just read about. Point the visual element inspector at anything on screen and ask for a targeted edit.
For a design mock-up rather than a working page, ask for a Design: a set of artboards on one canvas, each a screen, step in a flow, or poster. Select an element on an artboard to edit its text, colour, background or font size, and edits save automatically. Export any artboard as a PNG, or as a PDF through your browser's print dialog, or publish the whole canvas to a link.
Add knowledge to a Space by uploading files, fetching a link or pasting text, and the assistant can ground its answers in it. Connect a GitHub repository to browse its files, review the changes Studio proposes as diffs and commit them without leaving Studio.
Generate images and videos from a prompt on the Media screen. Image and video allowances depend on your plan.