Always Free · No Signup

Extract text from any image

Turn screenshots, scans, and photos into text you can copy. Best on clean printed type — handwriting is unreliable and busy multi-column layouts come out roughly. Your image is processed in your browser and never uploaded.

Runs entirely in your browser. Your files never leave your device.

Text language

Pick the language printed in the image, not your interface language. The language pack downloads once, then it is cached for next time.

How to extract text from an image

  1. 1

    Choose the text language

    Pick the language that appears in your image. This is set from your interface language by default, but a screenshot of an English page still needs English selected.

  2. 2

    Add your image

    Drop in a PNG, JPG, WebP, or HEIC file up to 20 MB. The first run takes a moment to set up the language you picked; the reading itself happens on your device.

  3. 3

    Copy or download the text

    Line breaks are preserved so paragraphs stay intact. Copy the text to your clipboard, or download it as a plain .txt file.

Image to text questions

Select the language printed in the image, then drop in a PNG, JPG, WebP, or HEIC file. The tool runs optical character recognition on your device and shows the result in an editable text box, with line breaks preserved. From there you can copy it to the clipboard or download it as a .txt file.

Not reliably. Text recognition of this kind is built for printed type, so neat block capitals sometimes come through while ordinary cursive usually does not. If your source is handwritten, expect to retype most of it — no free browser tool available today changes that.

English, Simplified and Traditional Chinese, Spanish, French, German, Japanese, Korean, Portuguese, Russian, Arabic, and Hindi. Choose one language per run: the engine loads a single trained model, so a page mixing two scripts is best done twice, once per language. Right-to-left output such as Arabic is displayed in the correct direction.

Accuracy depends almost entirely on the input. Low resolution, motion blur, JPEG artefacts, tilted pages, low contrast, and decorative fonts all cause character errors, and the confidence score above the result tells you when the engine itself is unsure. Complex layouts are the other common cause: multi-column pages, tables, and receipts often come out with the reading order scrambled even when individual characters are correct.

No. The reading happens inside your own tab, and the text stays in the page until you copy it or close it. We deliberately keep the recognised text out of our analytics too: a run reports the language you picked, the confidence score, and the character count — never the characters themselves. That matters here more than on most of these tools, because OCR is often pointed at receipts, contracts, and screenshots of private messages.

Aim for text that is at least about 20 pixels tall — roughly a 300 DPI scan of a printed page. Very large images are scaled down before processing because past around 2000 pixels accuracy stops improving while processing time keeps climbing. A sharp, well-lit, straight-on photo beats a high-megapixel photo taken at an angle every time.

More free tools

Background Remover

Upload a photo and get a clean transparent PNG in seconds. No signup, no watermark, no quality loss.

Try Now

AI Image Upscaler

Enlarge any photo 4× with AI. Rebuilds edges and texture instead of just stretching pixels.

Enhance

Depth Map Generator

Turn any photo into a depth map. Useful for design, 3D parallax effects, and creative compositing.

Generate

Object Detector

Spot and label objects, people, and items in any photo. See what's in your image with bounding boxes.

Detect
The Other Direction

Turn Text Into an Image

OCR reads what is already printed. The studio runs the opposite way — write a line and get a visual built around it: a slide, a quote card, a poster, a product label with your own wording set into the design.

  • Generate slides and quote cards from a line of text
  • Set your own wording inside a generated layout
  • Edit an existing graphic by describing the change
Explore Studio Features
The word Berlin built into a row of building facades, generated from a single line of text