iFood menu photo guide for restaurants

Accuracy matters more than polish

iFood's public guidance is especially useful for restaurants thinking about AI-generated or AI-enhanced menu images. The platform allows AI or illustrative imagery in defined circumstances, but the image still needs to avoid misleading customers about the product.

That makes iFood a good example of the broader delivery-photo rule: a better-looking image is not enough if it creates the wrong expectation. Menu visuals should help the customer choose, not promise an idealized version the kitchen will not serve.

• Use real photos whenever possible for high-trust menu items.
• Keep portion size, ingredients, and texture believable.
• Avoid artificial details that make the dish look like another product.
• Use clear disclosure when an image is illustrative or AI-generated.

Use AI as a workflow tool, not a disguise

AI enhancement can help restaurants correct weak lighting, background noise, and inconsistent framing. The wrong use case is turning a limited source photo into a fictional dish that does not match the product.

A practical iFood workflow should keep a human review step after enhancement. The reviewer should compare the output with the actual recipe, portion, packaging, and menu description before the image goes live.

• Enhance recoverable real photos first.
• Avoid generated ingredients that are not part of the dish.
• Reject images with unrealistic shine, texture, or serving size.
• Keep disclosure language aligned with local policy and platform rules.

Make the menu easier to maintain

iFood's menu tools emphasize structured product information, not only photos. That means the image workflow should be connected to item names, prices, descriptions, complements, and availability.

When a dish changes, the photo should be checked too. A menu refresh is weaker if the description changes but the image still shows an old recipe or discontinued side.