Creative update
Photo Analysis in Omni AI Companion
Photo Analysis gives a conversation a shared visual reference. Add a photo and ask a focused question about what is visible, whether you need a plain description, help organizing observations, or a second set of eyes on a design.
What changed
Questions now have a visual anchor
Instead of describing every visible detail yourself, you can place the image beside the question and discuss the parts that matter.
Observation and interpretation stay separate
A useful prompt can ask what is clearly visible first, then invite possibilities or suggestions. That distinction helps you make your own judgment about ambiguous details.
Why this feature matters
A photo can shorten the distance between noticing and explaining. A maker can ask for feedback on a poster's hierarchy, and a traveler can turn a street scene into a list of details to research later.
The feature should support attention, not replace expertise. It is a helpful companion for organizing what an image appears to show, but decisions about health, safety, identity, ownership, or important records deserve qualified human review.
Feature examples and use cases
- Design feedback: A volunteer shares a draft event flyer and asks which information is hardest to find at a glance.
- Everyday orientation: A traveler uploads a museum sign and asks for a transcription of the clearly readable parts.
- Visual inventory: A home organizer photographs a shelf and asks for a neutral list of visible categories before planning storage.
How to use it
- Add the image: Open Chat in Omni, use the attachment control, and choose a photo you have the right to share.
- Ask one useful question: State whether you want description, comparison, organization, or feedback. Point to the area of the photo that matters.
- Check the answer against the image: Zoom in yourself and ask a follow-up when something is unclear. Keep the final decision with your own review.
Tips for a better first try
- Ask for observations before asking for conclusions.
- Mention whether tiny text, colors, or spatial relationships matter.
- Crop out unrelated private details when possible.
Limitations and things to know
- Blur, glare, low resolution, or unusual angles can make details hard to interpret.
- An image-based response is not a diagnosis, identity confirmation, legal determination, or safety inspection.
Availability
Available in Chat when photo attachments and visual analysis are shown for your account. Open the attachment control to see the options currently available to you.
Frequently asked questions
Can it read text?
It can discuss text that is clearly visible, but you should verify every important transcription against the original.
Can I ask about several photos?
If the current chat accepts them, describe how the images relate and ask for a comparison; results still depend on clarity and context.
Should I upload sensitive documents?
Only share material you are comfortable discussing in the service, and crop or remove unnecessary personal details first.
Related Omni updates
Browse every Omni feature update ยท Explore step-by-step tutorials
