File uploads, live video search, etc.

Google is expanding its AI mode with new tools including PDF uploads, persistent planning documents and real-time video assistance.

Updates begin to unfold today, with the AI mode button appearing on the Google homepage for desktop users.

PDF uploads are now supported on desktops

Desktop users can now upload images directly to search queries, a feature previously only available on mobile.

PDFS support is coming several weeks away, so you can ask questions about uploaded files and receive AI-generated responses based on both the document content and related web results.

For example, students can upload lecture slides and use AI mode to gain an understanding of the material. The answer includes suggested links for deeper exploration.

Google will support additional file types and will integrate with Google Drive “in the coming months.”

Canvas: Tools for Multi-session Planning

A new AI mode feature called Canvas can be left organized across multiple search sessions.

When you ask AI mode for help with planning or creating something, you will receive the “Create Canvas” option. This opens a dynamic side panel that you want to save and update as your queries evolve.

Use cases include building study guides, travel itinerary, or task checklists.

Canvas is available for US desktop users registered in the AI Mode Lab experiment.

Real-time support for live search

This week we'll be searching live with video input, which will be released on mobile. This allows you to use AI mode while pointing the phone camera on real objects or scenes.

This feature is built on Project Astra and is available from Google Lens. Start by tapping the “live” icon in the Google app and use live video as visual context to have a before and after conversation with AI mode.

Chrome adds context AI answers

The lens extends desktop functionality within Chrome. You will immediately see the “Ask Google about this page” option in the address bar.

When selected, a panel will open where you can highlight parts of the page, such as text diagrams and snippets, and receive an AI overview.

This update also allows follow-up questions by directly selecting the button labeled “Dive Deeper” or the AI mode via AI mode from within the lens experience.

Looking ahead

These updates reflect Google's vision of searching as a multimodal, interactive experience rather than a one-off text query.

While most of these tools are currently limited to U.S. lab users, it points to a future in which AI modes will become the centre of searchers' exploration, learning, and planning.

The rollout timeline varies depending on the feature. So, be aware of how these features are added to your search experience and consider how to adapt your content strategy accordingly.