AI Image Generation

A built-in service that turns a written description into an image, or edits a photo the user picks. Ask for "an avatar generator", "a logo maker" or "an app that enhances my photos" in chat and GenMB wires it into your app. It works the same in the live preview and in your deployed app.

Generated apps call the SDK that GenMB injects. It resolves to a data URL you can put straight into an image tag:

const src = await window.genmb.ai.generateImage('a friendly fox mascot, flat illustration', {
  size: '1024x1024',   // or '1024x1536' (portrait), '1536x1024' (landscape)
  quality: 'medium',   // or 'low'
  transparent: false,  // true returns a PNG with no background
})
img.src = src

To edit a photo, pass the file (from a file input) or its URL with an instruction. The result keeps the photo's shape unless you pass a size:

const src = await window.genmb.ai.editImage(file, 'brighten it and remove the background')

• 1 credit per image, generated or edited, charged to the app owner, only when an image is returned.
• Photos to edit can be PNG, JPEG or WebP up to 10MB; the SDK shrinks larger photos and converts other formats first.
• Per app: 20 images per hour, 100 per day, and 2 at the same time. Over the cap the SDK throws an error with code RATE_LIMITED.
• When the owner is out of credits the SDK throws an error with code INSUFFICIENT_CREDITS.
• Prompts up to 4,000 characters.

FAQs