Processed on our server. Your input is sent to our server to produce the result. Uploaded files are not kept, and nothing is sold or shared; our privacy policy lists what is logged.
Translate Text From a Photo
The Tools Kit's free Photo Translator turns a picture of English text into a translation in another language. Snap a sign, menu, product label, notice, or screenshot, upload it, and the tool reads the words out of the image with OCR (optical character recognition), lets you correct them, and translates them into Spanish, French, German, Italian, Portuguese, Hindi, Urdu, Arabic, Chinese (Simplified), Japanese, or Russian.
It works in two stages. The text recognition runs inside your browser, so the photo itself never leaves your device. Only the extracted text, up to 500 characters, is sent for translation when you click Translate.
How to Use the Photo Translator
- 1
Upload a photo with English text
Click the upload box or drag an image onto it. JPG, PNG, and WebP photos, screenshots, and scans all work.
- 2
Let the OCR read the text
Text recognition runs in your browser and shows a progress percentage. The recognized English text then appears in an editable box.
- 3
Fix any misread words
Correct OCR mistakes, delete stray symbols, or trim the text to the part you need. The first 500 characters are what gets translated.
- 4
Choose a language and click Translate
Pick a target language (Spanish, French, German, Italian, Portuguese, Hindi, Urdu, Arabic, Chinese (Simplified), Japanese, and Russian) and click Translate. Copy the result with one click, or click Start Over to translate another photo.
If you only need the English text and no translation, the Image to Text converter runs the same in-browser OCR without the translation step.
How the Image Translator Works
Step one: OCR in your browser. The tool loads Tesseract.js, an open-source OCR engine, along with its English language data. It scans the image for letter shapes and returns the text it recognizes while a progress percentage counts up. Because this happens on your device, a photo of a document or a private screenshot is never uploaded to our server.
Step two: machine translation. When you click Translate, the text in the editable box is sent to our translation endpoint, which forwards it to the MyMemory translation service from English into the language you picked. The result appears in a read-only box with a Copy Translation button.
The recognition model is English only, so this is a one-way English-to-other-language picture translator. It cannot translate a photo of Japanese, Arabic, or Spanish text into English. It also returns plain text rather than drawing the translation back onto the image.
Get Better Results From Photos
Translation quality depends almost entirely on how well the OCR reads the photo. A misread word gets translated as the wrong word, so a few seconds spent on the image pays off.
- Shoot straight on, in good light. Angled shots, glare, and shadows make letters harder to recognize. If the photo came out sideways, fix it with the Image Rotator before uploading.
- Crop to the text. Busy backgrounds, logos, and pictures produce stray symbols. Use the Image Cropper to keep just the paragraph or menu section you want to translate.
- Use a supported format. JPG, PNG, and WebP read reliably. iPhone photos saved as HEIC may fail, so run them through the HEIC to JPG converter first.
- Prefer printed text. Tesseract handles printed and typed text well; handwriting and decorative fonts are often garbled.
- Edit before you translate. Fix misread letters, delete stray characters, and join lines that were split mid-sentence, since broken sentences translate poorly.
Translating Longer Text: the 500-Character Limit
Each translation covers the first 500 characters of the text in the editable box, including spaces and line breaks. Anything after that point is not translated, so for a long page you translate it in sections: copy the result, delete the part you have translated from the box, and click Translate again.
For example, a photographed notice with 1,200 characters of text takes three passes: 500 + 500 + 200 characters. To see how long the extracted text is, paste it into the Word Counter, which shows the character count, and collect the translated sections in the Online Notepad as you go.
Common uses
- Translating an English menu, street sign, or notice for a friend or family member who reads another language
- Turning a screenshot of English instructions or an error message into Spanish, Hindi, Urdu, or Arabic
- Translating a product label, ingredient list, or packaging text from a photo
- Getting the gist of a printed handout or letter in your own language
- Pulling text from a photographed page to check or reuse it, then cleaning it up in the Grammar Checker before translating
More text tools, from summarizing to citations, are collected in the writing tools section of The Tools Kit.
Photo translator terms explained
| Term | What it means here |
|---|---|
| OCR (optical character recognition) | Software that finds letters in an image and turns them into editable text; here it is Tesseract.js running in your browser |
| Source language | The language of the text in the photo; this tool reads English only |
| Target language | The language you want the text translated into; you can choose from 11 |
| Machine translation | Automatic translation by software rather than a human translator; fast and useful for the gist, but not certified |
| Character limit | The maximum amount of text sent per translation; here, the first 500 characters of the extracted text |
| Contrast | How strongly the text stands out from its background; dark text on a plain light background reads best |
Frequently Asked Questions
Upload the photo, wait for the OCR step to read the English text, correct any misread words in the editable box, then choose a target language and click Translate. The tool supports Spanish, French, German, Italian, Portuguese, Hindi, Urdu, Arabic, Chinese (Simplified), Japanese, and Russian.
Yes. There is no signup, no account, and no payment. You work on one image at a time, and there is no per-photo charge for normal use.
No. Text recognition runs in your browser with Tesseract.js, so the image itself is never sent to our server. When you click Translate, only the extracted text (up to 500 characters) is sent to our translation endpoint, which passes it to the MyMemory translation service.
English only. The OCR step uses an English recognition model and there is no source-language option, so photos of Spanish, Chinese, Arabic, or other non-English text will not be read correctly. The tool translates from English into another language, not the other way around.
Spanish, French, German, Italian, Portuguese, Hindi, Urdu, Arabic, Chinese (Simplified), Japanese, and Russian. That is 11 target languages in total.
Yes. Each translation covers the first 500 characters of the text in the box; anything beyond that is not translated. For a longer passage, delete what you have already translated and click Translate again for the next section.
Most errors start in the OCR step. Blurry, tilted, low-light, low-contrast, stylized, or handwritten text is often misread, and the translator then translates the mistakes. Retake the photo straight on in good light, or fix the misread words in the editable box before translating.
Only poorly. Tesseract is built for printed text, so neat block capitals may partly work but cursive and messy handwriting usually come out garbled. Typing the text into the box yourself is often faster in that case.
Convert it to JPG first. The OCR engine reads common formats such as JPG, PNG, and WebP, and a HEIC file may fail to read in most browsers. Switching the iPhone camera to Most Compatible also saves new photos as JPG.
No. The tool returns plain text: the extracted English in one box and the translation in another, with a Copy Translation button. It does not draw the translated text back onto the photo.
No. It is machine translation, which is good for understanding a sign, menu, label, or short notice but is not a certified translation. Use a professional translator for legal, medical, immigration, or other official paperwork.
Reviewed by Raja Jahangir · Last reviewed: October 7, 2026
How do you translate text from a photo?
To translate text in a photo, upload a JPG, PNG, or WebP image; English text is read locally in the browser with Tesseract OCR. Edit any misread words, choose a target language such as Spanish, French, German, Hindi, Urdu, Arabic, Chinese, or Japanese, and click Translate. Translation sends up to 500 characters to a translation service.
Why is the text from my photo translated wrong?
Most errors come from the OCR step, not the translation. Blurry, angled, low-contrast, or handwritten text is misread, and the recognition model is English-only, so non-English source text is not extracted properly. Retake the photo straight-on in good light, then correct any misread words in the editable box before translating.