Turn any web image or screenshot into the AI prompt that made it — right-click or capture, copy a ready-to-use prompt in seconds.
Right-click any image and choose "Get prompt" — or click the toolbar icon to screenshot any region of the page.
Preview exactly what you're sending, then generate. Takes a couple of seconds.
Drop the finished prompt into Midjourney, Stable Diffusion, DALL·E, or any image model.
Reverse-engineer a look you love instead of guessing keywords. Works on any image rendered on the page.
Drag to select any region — video frames, canvas renders, PDF previews, protected images. If you can see it, you can capture it.
Every prompt you generate is saved locally with its source image and page — your last 30, ready to revisit without spending another credit.
Drop images onto a private board as you browse — no sign-in, no credit spent. Arrange them freely, select several at once, and reverse only the ones worth a prompt.
Try it before you commit — no card required.
200 generations a month. Cancel anytime.