How to Ask the AI to Analyze an Image You Send It
What this feature does
Beyond generating text, SoulrisAI's chat can look at an image you send it and answer questions about it — describe what's in it, read text within it, help you understand a chart or screenshot, or give feedback on a photo. This is different from image generation (asking the AI to create an image for you) — this is the AI reading an image you already have.
What you need before you can use it
Image analysis in chat requires a funded wallet. It isn't available on the free tier — see why attachments need a deposit for why. Once your wallet has a balance, the option to attach an image appears in the chat interface.
How to send an image
In your chat, use the attachment option to select an image from your device, then add your message alongside it — for example, attach a photo and ask "what's wrong with this plant" or attach a screenshot and ask "explain what this error means." You can ask your question in the same message as the attachment, or attach the image first and ask follow-up questions about it afterward in the same conversation.
Supported image types
JPEG, PNG, and WebP images are supported (see supported attachment file types for everything chat accepts). If your image is in a different format, convert it first — most phone cameras and screenshot tools already save in one of these.
Which model handles it
Image analysis requires a model with vision capability. SoulrisAI routes this automatically when you attach an image — you don't need to manually pick a vision-capable model, though you still can if you prefer a specific one (see choosing a model manually).
Does it remember the image later in the conversation?
Yes, within the same conversation thread — once you've sent an image, you can keep asking about it in follow-up messages without re-attaching it. Starting a new conversation, though, means the AI no longer has it — see how conversation memory works.
How it's billed
Attachments are billed as part of the message cost, on top of the usual chat usage — analyzing an image adds to what that message costs compared to a text-only message. There's no separate flat fee just for the feature; cost scales with the message it's part of.
Good use cases to try
- Getting help understanding a chart, diagram, or screenshot
- Asking for feedback on a photo (composition, what could be improved)
- Reading text out of an image when you can't easily copy-paste it
- Troubleshooting something visual — an error message, a broken layout, a diagram that doesn't make sense
If what you actually want is the AI to create an image rather than analyze one you already have, that's covered separately — see how to generate your first AI image once you're ready for that instead.