Palimpix

Ctrl+click any image to translate the text in it with AI (Gemini, ChatGPT or Claude, your own key). Shown in place.

Translate the text inside any image on any website, right where it is. Hold Ctrl (Cmd on a Mac) and click an image: Palimpix reads every piece of text in it with AI, erases the original text and draws the translation in its place, like a scanlation. Works on manga, comics, memes, screenshots, posters, infographics and product photos. FEATURES • Use your own AI key: Google Gemini (has a free tier), OpenAI (ChatGPT) or Anthropic (Claude). • Auto mode picks a suitable model from your key's model list and switches to another model when one is busy. • Translate into Arabic (default) or about 30 other languages. Right-to-left languages such as Arabic, Persian, Urdu and Hebrew are laid out right to left. • Clean results: on speech bubbles and flat backgrounds only the letters are removed and the bubble outline is kept; on gradients and photos the background is blended. • Works on normal images, images from other sites, hotlink-protected images, CSS backgrounds, canvas, video frames, X/Twitter photos and very tall webtoon strips. • Hover the image for Show original, Copy text and Close. Ctrl+click again to switch between the translation and the original. • Right-click any image and choose "Translate text in this image". • Clear messages for a wrong key, a used-up quota, a busy model or a blocked image, with a seconds counter while it works. • Choose your shortcut: Ctrl, Alt, Shift or Ctrl+Shift + click. Normal clicks are never touched. HOW TO START 1. Get an API key from Google AI Studio, OpenAI or Anthropic. 2. Paste it in Palimpix settings and press Save. 3. Ctrl+click any image. PRIVACY • Your API keys are saved only in this browser (never synced) and are sent only to the provider's official API. • Only the image you click is sent, and only to the provider you chose. • No Palimpix server, no analytics, no tracking. • Open source (MIT): https://github.com/Niro54321/Palimpix