Image Prompting
You can include images in your prompts to give the AI a visual reference. This is useful for recreating designs, reporting layout bugs, matching a brand style, or starting from a wireframe or mockup.
Adding Images to a Prompt
Section titled “Adding Images to a Prompt”Drag and drop — drag a PNG or JPEG file directly onto the message input.
Media Library — click the media button near the prompt input, browse your uploaded images or search Unsplash, then click Add to Prompt.
Images should be PNG or JPEG, under 10MB, and as clear as possible — the AI reads them literally, so blurry or low-res images produce less accurate results.
What to Use Images For
Section titled “What to Use Images For”- Design recreation — upload a Figma export or screenshot and ask the AI to build it
- Visual debugging — screenshot a layout issue and describe what’s wrong
- Brand references — share a logo, color palette, or style guide to keep output consistent
- Wireframes — sketch a rough layout and let the AI fill in the styling
Best Practices
Section titled “Best Practices”Pair every image with a clear text instruction that tells the AI what to do with it. The more specific, the better:
- Good: “Recreate this navbar with the same spacing and colors”
- Vague: “Make something like this”
You can also include multiple images in one prompt — useful for before/after comparisons or referencing different parts of a design system.