Can antigravity support native multimodal reading of PDFs, like in AIstudio and Gemini Cli?

Gemini API
,
1

This feature will be very useful, reading some documents and papers.

3

Hello @yi_ren, Google Antigravity supports native multimodal reading of PDFs. In Antigravity, PDF support is integrated directly into the workspace file system. This allows the agent to treat PDFs as part of your project’s codebase or documentation. The feature provides multimodal reasoning and structured extraction.
To enable an agent to read a PDF in Antigravity, you could follow these steps:

  1. Upload to Workspace: The PDF must be uploaded to your project’s file tree (the “Explorer” view).

  2. Trusted Folders: You must place the PDF in a folder that has trusted permissions. Antigravity uses a security model where agents cannot read files in untrusted directories by default. Once the folder is trusted, the agent can “see” and analyze the document.

  3. Agent Interaction: You can then reference the file in your natural language prompts.