Camera
Show it instead of describing it
Hold a device, a page, or a physical setup up to the camera and ask about it. The assistant follows what is in frame and answers about what it can actually see.
Vision
Live vision AI that reads the room with you
Point a camera at what you are working on, or share the window you are stuck in. Live vision keeps the picture inside the conversation instead of asking you to describe it in words.
What you get in a live session
Screen
Read the window you are in
Share a window and ask about the chart, the error message, or the paragraph you are reading, without copying anything into a chat box.
Follow-through
Guidance that moves with you
When the screen or the scene changes, the advice changes with it: correct the step you just took instead of repeating the whole list.
Frequently asked questions
What does live vision AI mean here?
It means the assistant can take in a continuous visual feed, from a camera or a shared screen, while the conversation is happening, rather than analysing one uploaded image at a time.
What can it actually see?
Only what you turn on. With the camera you decide what is in frame, and with screen share you choose the window or tab. Either one can be stopped at any moment.
Do I have to upload files?
No. Camera and screen share replace the upload step during a live session. Uploads still work whenever a single still frame is all you need.
How accurate is it?
It is a general model reading a live feed, so treat it as a fast first opinion. It is strong at reading text, layout, and obvious problems, and worth double-checking for precise measurements or anything safety critical.
When should I use screen share instead of the camera?
Use screen share for software, documents, and dashboards. Use the camera for physical objects, printed pages, or anything sitting on a desk in front of you.
See it for yourself
Share a screen or point a camera and ask the question you actually have, not the one you can put into words.