Read text out of images in 20+ languages, ready to translate. Our image translator uses multilingual OCR to pull foreign-language words from a photo or screenshot so you can translate them in seconds.
What is the Image Translator?
The Image Translator is a free online tool that reads foreign-language text out of an image so you can translate it. It uses multilingual OCR, optical character recognition tuned for many languages and scripts, to recognise the characters in a photo, screenshot or scan. You upload the image, the tool extracts the text in its original language, and you copy that clean text into the translator of your choice.
The key idea is separating reading from translating. A photo of a foreign menu, a sign, or a page in another language is just pixels until something reads it. Getting an accurate transcription of the source first is what makes the eventual translation trustworthy, because a mistranslation often starts as a misread character. If you only need English printed text, Image to Text is the simpler tool. Our guide on translating text from an image walks through the full two-step flow.
How to extract text from a foreign-language image
- Upload your image by dragging it in or browsing. JPG, PNG, WebP and other common formats work.
- Choose or confirm the source language so the engine knows which characters to expect.
- Let the OCR engine read the text and produce a transcription.
- Download or copy the recognised text, then paste it into your preferred translation service.
No install, no account. Selecting the right language up front is the single biggest thing you can do to improve accuracy.
Where it helps
Multilingual capture shows up constantly when language is a barrier. Travellers read menus, signs and tickets. Students work through a passage in a language they are learning. Importers and researchers extract text from foreign documents and labels. In each case the picture has to become text before any translation, dictionary lookup or search can happen.
Because the output is plain, editable text in the original language, it slots straight into a translator, a dictionary, or a document. If you have many foreign-language images to get through, Batch Image to Text processes a whole set at once, and if your source is handwritten, Handwriting to Text is the better starting point.
Getting accurate multilingual OCR
Pick the right language
The engine reads more accurately when it knows which script and language to expect, because that narrows the set of possible characters. Always set the source language rather than leaving it to guess, especially for accented or non-Latin text.
Capture cleanly
Photograph the text straight-on in bright, even light, and fill the frame. Non-Latin scripts and diacritics depend on resolution to be read correctly, so a higher-resolution capture matters even more here than for plain English. The capture advice in our 12 ways to improve OCR accuracy applies directly.
Mind the script direction
Right-to-left scripts like Arabic and Hebrew, and dense CJK characters, are more sensitive to skew and low resolution. Keep the page flat and the camera square. A clean source image prevents most of the errors that would otherwise surface only after translation.
Read, then translate
This tool deliberately focuses on the reading step. Doing it well is what separates a good translation from a garbled one, because translation tools can only work with the text they are given. Once you have clean source text, paste it into whichever translator you trust for a final, readable result.
Honest expectations
The Image Translator uses a Tesseract-based OCR engine with multilingual support. It is strong on clean printed text across many languages and scripts, and best effort on stylised, low-resolution or handwritten sources. No OCR reads every script perfectly, so review the extracted text before translating, particularly accents and unfamiliar characters. For most clear printed images, you will have accurate, translation-ready text in seconds. Upload an image to the Image Translator above to start.