A little guidance goes a long way.

Get comfortable with your connected workspace.

Start with a question

Type a prompt in Home or Chat. Attach context with +, use voice dictation, and choose a response profile in Settings.

Open chat

Bring your documents

PDF, DOCX, text, CSV, JSON, and Markdown are parsed in your browser. Scanned pages need Vision image analysis.

Work with documents

Find anything quickly

Press ⌘ K or Ctrl K to navigate or search titles across your conversations, projects, and saved library.

Browse library

Understand your privacy

Account records are private. Uploaded files are public. AI processes the text or images you explicitly submit.

Review privacy

Connected capabilities & limits

Available: Google and email/password authentication, private key-value storage, AI text completion, image generation and editing, image analysis, and public file storage.

Live research: The dedicated Web Search capability requires a Pro plan. Upgrade and enable it from the Add-ons panel. Until then, Research works with the sources you provide and does not claim live verification.

AI transport: The built-in service returns complete responses, not token streams. Stop dismisses a pending result but cannot cancel the provider request. Direct Gemini model configuration is not exposed by the enabled SDK.

Code: HTML/CSS/JavaScript previews are isolated with network access disabled. No uploaded or generated code executes on the server.

Storage: Private records use the built-in key-value store. A relational database is not connected. You can connect Supabase from the Database panel if your project needs relational data.