Image to Text (OCR)

Retyping text from an image is exactly as tedious as it sounds — and entirely avoidable. This tool performs OCR (optical character recognition) in your browser, powered by the Tesseract engine running via WebAssembly. Load an image — a screenshot, a photo of a printed page, a scan of a receipt — choose the language, and get the text out as plain, copyable, editable text.

What it is good for: pulling a paragraph out of a screenshot instead of retyping it, digitizing a printed form or letter, extracting the total from a receipt photo, recovering text from an image-only PDF page, or grabbing a code snippet from a conference slide. Multi-language documents are supported: you can select several recognition languages at once, which helps when a page mixes, say, English text with Chinese annotations.

Two honest expectations to set. First, the language model downloads to your browser on first use, so the initial recognition takes a moment; afterwards the model is cached and subsequent runs start immediately. Second, accuracy depends heavily on image quality. Crisp, high-contrast, straight-on images of printed text recognize very well; blurry photos, curved pages, decorative fonts, handwriting, and low-light shots degrade results noticeably. If output looks garbled, try a sharper, brighter, more level photo before anything else.

Everything runs on your device: the image is analyzed locally and never uploaded to a server, which makes the tool usable for screenshots containing credentials, personal correspondence, or internal documents you would not paste into an online OCR service. After recognition, read the output before using it — OCR is never perfect, and a quick proofread catches the stray character a spellchecker would happily ignore.