For questions regarding recipe clipping, please see the below per-device instructions. For all other questions, please see our FAQ, or contact support.

• Support Index
• • iOS Recipe Clipping
  • Photo Recipe Import
  • Text Recipe Import
  • MacOS Recipe Clipping (Select Below)
  • Safari Browser Extension
  • Share Extension

• Recipe Photo Import (iOS)
• Return to top

  

  • To import a photo on iOS or iPad: Take a high-quality photo of the recipe in the brightest possible light, ensuring recipe text is crisp and clear. For longer recipes, Cookery supports up to two photos per recipe, a limitation of Apple Intelligence support. Tap the button in Cookery to open the New Recipe menu and select Import from Photo, if available.

    

  If Recipe Photo Import is unavailable, verify that your device has updated to iOS 26, and that your device supports Apple Intelligence.

  Click here for Apple Intelligence device requirements.

  • Choose up to two photos for the recipe (Apple Intelligence has a limit on the amount of text it will support), then tap Convert.

    The recipe conversion process will begin, filling out the recipe as the AI categorizes the text. Once this completes, you can edit the values before saving. Alternately, if there are too many mistakes, you can try the to attempt to improve the recognition process, or see tips below.

  • If you notice multiple errors in your recipe result, this can often be a photographic issue:

    • When taking photos of a recipe document, use non-digital zoom (we have found 2x zoom to be effective) or the camera's distance from the recipe to get as close to the recipe as you can while ensuring focus and sharpness remains high.
    • Ensure lighting for your photo is very bright, with as few shadows as possible.
    • If taking a photo from a recipe book, try to flatten the book so the text isn't warped by the binding, and so the binding doesn't cause a shadow on the desired page.
    • When using OCR to read text in photos, it is essential to ensure (if possible) there is no text in the image that is not a part of the recipe (this can include page numbers if from a cookbook, etc.). If the recipe contains text that isn't a part of the recipe title, servings/yield, description, ingredients, or instructions, this may result in a malformed recipe. If possible, try cropping unneeded text out of the photo.
    • Try using the available on the recipe, as this can help improve OCR image-to-text readability.
    • If you've tried the above suggestions and the AI still fails to categorize the text correctly, try running it again -- each attempt is a separate session with the model, and will probably result in subtle output differences.
    • Some characters that are used in recipes, such as ½,¾, or other measurement abbreviations, can be difficult for optical character recognition to discern. Double check numerical measurements in the original recipe (especially those with uncommon characters).
    • If you're still seeing problems with recipe formation, please contact us.