Snap a photo and let AI fill in the details

When you add an item, you can take or pick a photo and let AI read what it can off the image. Manifest fills in the fields for you to review — you fix anything wrong before saving.

How to use it

1. Tap the add button and choose to add a single item.
2. On the photo step, tap Take a photo to use your camera or Choose from library to pick an existing image.
3. Manifest sends the photo to AI. Manifest fills in what it finds.
4. Review the pre-filled fields, correct anything that looks wrong, then save.

What gets filled in

• Name, brand, and model
• Serial number, UPC, and ISBN — read directly from visible labels, barcodes, or packaging
• Category

If the photo is a receipt or invoice, Manifest also reads:

• Retailer name, price paid, and purchase date

Each field shows a confidence level (high, medium, or low) so you can see how sure the AI was. Low-confidence fields are worth double-checking.

Also detected

If the photo is clearly a pet with no product in frame, Manifest switches to pet mode instead of filling in product fields — and does not charge an AI action for it.

If the photo shows a gift card, Manifest switches to gift card mode and reads the card's retailer, face value, and code.

Good to know

• Analyzing a photo uses one AI action.
• The AI reads text it can see. A blurry label or poor lighting will produce lower confidence results.
• If the result is too far off to be useful, you can re-take the photo or fill in the fields yourself.
• Photos of receipts and invoices also work here, but uploading a document file (PDF or image) is covered in Scan a receipt or PDF in the app.