Skip to main content
You can provide context to the assistant by attaching files, images, screenshots, or working with the current browser page.

​
Attachment menu

Click the attachment icon in the composer toolbar to open the menu. You can add:
  • File — Upload a document for the assistant to read.
  • Image — Upload an image for analysis.
  • Screenshot — Capture the current browser page.
  • Model selector — Switch your active model.

​
File limits

  • Maximum file size: 20 MB per file.
  • Inline text budget: 12,000 characters per attachment, 24,000 total characters across all inline chat attachments.
  • Files exceeding these limits will show an error.

​
Supported file types

Documents: txt, md, markdown, json, csv, js, jsx, ts, tsx, py, html, htm, css, xml, yml, yaml, sql, log, pdf Images: png, jpeg, gif, webp

​
Screenshots

Select Screenshot from the attachment menu. You can capture the visible portion of the current browser tab. The assistant can then answer questions about what’s on the page.

​
Active resources

When you add files, images, or other context, they appear in the Active Resources bar above the composer. You can review what’s attached at a glance.

​
Context panel

Open the Context view from the header to review and remove attachments, or manage long-term memory.