File Translator
Translate small PDF, DOCX or photo text. The text obtained is sent to online AI.
File TranslatorFile Translator
Extract text from small Word (.docx), text or scanned PDF, and JPG/PNG/WebP photos. There will be machine translation, check it in the required documents.Translate text in a small Word (.docx) file, a PDF, or a JPG, PNG or WebP image. Scanned PDFs and photos use OCR. Please check the result before using it for anything important.
One file maximum 2 MB, PDF maximum 3 pages and total 2,200 characters in translation. DOCX attempts to save paragraphs, tables, and images, but changing words will not preserve the exact design. PDF/Photo will create a new PDF with a simpler design; There may be errors in OCR on scanned PDF.Limit: one file up to 2 MB, a PDF of up to 3 pages, and up to 2,200 characters of extracted text. A translated DOCX keeps its original structure and styling where possible, but its layout may change. A PDF or photo produces a new PDF with a simple layout. OCR can make mistakes.
The file is read in your browser. Only the extracted text is sent to Cloudflare Workers AI for translation, not the file. OCR of photos/scans is done in the browser with Tesseract; Data of his/her language can be downloaded. Turnstile bot check is implemented. Do not upload sensitive files.Your browser reads the file locally. Only the extracted text, not the file, is sent to Cloudflare Workers AI for translation. Tesseract runs OCR in your browser and may download language data. Cloudflare Turnstile checks for bots. Do not use sensitive files.
See TranslationReview translation
DOCX to DOCX will be downloaded. New PDF from PDF/Photo; Its layout will not be the same as the original.A DOCX downloads as a DOCX. A PDF or photo downloads as a newly laid-out PDF, not a copy of the original layout.
Also open Text/Voice Translator
The mic will listen in the language selected in "From what language" above, then translate into the language selected in "What language from." Requires microphone permission in browser. Not every browser recognizes the sounds of every language; If this happens then write the text and translate it.
Note: Up to 3000 characters at a time. This is machine translation (mainly Google Translate, if not MyMemory), so please check it yourself before using it in important documents.
How to use File Translator?
The job of File Translator is to extract text from PDF, DOCX or image and translate it into another language. This facility is provided in a separate page on PDF Saathi so that you can read the correct input, output and required limits. The method written below is according to the actual tool's behavior, there is no guarantee of perfect conversion or approval.
Steps and output
Choose a small file, set the source and target language and translate once the bot check is complete. Check result preview and download. Result: translated DOCX or new PDF. Set aside a safe copy of the original file before starting work. Selecting the wrong page or version may require rework, so match the file name and content first. Start with a small sample first. Do not close the tab while the tool is running. If you see an error, read it; Do not consider getting incomplete results as success.
Where is caution necessary?
Existing limits: 2 MB file, PDF maximum of 3 pages and 2,200 total extracted characters. The extracted text is sent to Cloudflare AI. Output layout is not guaranteed to be original. After downloading, open the new file in a separate viewer. Match page count, orientation, small text, names, dates and important numbers with the original document. Mere downloading of the file is not proof of correct result. If the output is to be sent to a government form, office, school or customer, also check the current file-size and format requirements of that place. If important information has been cut out, appears too light or is misplaced, recreate it from the original.
Privacy and safe use
This is online translation or AI facility. External service may be used for text processing, so do not consider it as offline or completely local feature. Do not enter personal IDs, passwords and private account information. Do not forward the machine output without reading it. Start with a normal document and sample text, then check the result according to your needs. Work only after reading the limits given on this page. Browser memory is different on every device, so no fixed speed or guaranteed result can be given for very large files. Download the new file after processing. Complete the test before replacing the original. Change someone else's document only if you have their permission.
Frequently asked questions
Is the file uploaded to the server?
The file is read locally, but the extracted text is sent for AI translation. Do not include sensitive files.
Will the original file be changed?
No. Save the new output copy separately. Keep the original safe and read it before sending the final result.
What if an error occurs on a large file?
Try with a smaller sample or fewer pages. If memory is less then run it on another device. Do not use incomplete output after error.