guides · 2026-10-30 · 1 min read
How to copy text from an image (free OCR in your browser)
ocr (optical character recognition) turns the text in a picture into text you can copy, edit and search. it is handy for a quote in a screenshot, a page of a printed book, a receipt, a whiteboard photo or a scanned letter you need to reply to.
get the text out
- open image to text.
- choose the language of the text (english is the default; there are over twenty).
- drop a photo, screenshot or scanned pdf on the box.
- wait while the engine loads and reads the text. the first time, it downloads the language data once.
- fix any mistakes in the box, then copy the text or download it as a txt file.
everything runs in your browser using tesseract, an open-source ocr engine. your image is never uploaded, which matters for ids, letters and receipts.
get better results
- light and focus: a sharp, evenly lit photo is the biggest improvement. avoid shadows from your hand or phone.
- straight on: hold the camera parallel to the page. if it is tilted, straighten it with rotate image first.
- crop: cut away everything that is not text with crop image, especially pictures and patterned backgrounds.
- size: tiny text in a screenshot reads better after you enlarge it with resize image (200% often helps).
- language: pick the right one; reading spanish text with the english setting gets the accents wrong.
what ocr struggles with
handwriting (results vary a lot), unusual or decorative fonts, text over photos, and complex layouts like tables and multi-column pages. for tables, expect to fix spacing by hand. always proofread numbers, names and amounts before you rely on them.
other languages and screenshots
there are ready-made pages for spanish, hindi, chinese and text from a screenshot. screenshots usually give the cleanest results because the text is perfectly sharp.
after you have the text
paste it anywhere, check its length with the word counter, or fix capitalization with case converter.