Working with images

Assistants with vision can look at images you share and answer questions about them, useful for screenshots, diagrams, photos of documents, and charts.

Share an image

1. In the message box, upload an image file, or simply paste one from your clipboard.

2. Add as many as you need to a single message.

3. Ask your question, “what does this error mean?”, “summarise this chart”, “read the handwriting on this form”.

Reading documents with vision

Ajutant can also read scanned or handwritten documents using vision rather than plain text extraction, which handles messy scans far better. When you upload such a document, the assistant can work from what it sees on the page.

Note

Vision is a per-assistant capability. If an assistant can’t accept images, you’ll be told when you try to attach one.