under.pics

Right-click any image on the web to read the prompt behind it, then generate your own from it.

under.pics - read the prompt behind any image, then make your own. Right-click an image anywhere on the web, or use the button that appears when you hover it, and pick "Remake this image". A panel opens over that picture with thirteen things you can do to it, grouped so the strip stays two rows: - Remake - writes one detailed prompt describing the subject, composition, lighting, palette, materials and mood, then paints a new image from it - Face Swap - puts the face from a picture you attach onto the body in this one - Character Swap - replaces the whole person, keeping the place and the light - Makeup - reads the face, then does the makeup that suits it - Turn into Anime and Turn into Real - the same picture, each way round - Turn into Watercolour - repaints it as a watercolour on white paper - Sketchbook Page - redraws it as a hand-drawn sketch on paper, with a title and a caption written beside it - Postage Stamp - prints it onto a perforated stamp, in its own colours, under a postmark - or switch to Engraved and it is redrawn in one ink as ruled lines - Fridge Magnet - cuts it out of clear acrylic and sticks it on the fridge door - Turn into Sticker - cuts the subject out and draws it as a die-cut sticker - Remove background - Upscale ×4 Nothing leaves your browser until you ask for it. Hovering sends nothing. Opening the panel sends nothing. The image goes out on the press that starts the job - and the panel says so, on screen, above the button that does it. The prompt behind a generated image stays on screen with it, one click from your clipboard. The toolbar popup keeps your recent images one click away too. You bring your own backend: under.pics talks to one API, with one key you paste in Settings. It never holds provider credentials itself, and it cannot call any server until you grant that one origin on the click that saves your key.