Page image OCR · PNG / JPEG / WebP · Paired review
Translate scanned PDF content with OCR and image translation
A scanned PDF often contains page images rather than readable text. Export the pages you need as PNG, JPEG or WebP, then use Upload Document below. The file type is detected automatically and image recognition settings appear. PDFs with a usable text layer can be uploaded directly.
For image-only scans, export the required pages in your PDF reader and use Upload Document below, one image at a time. This is not an automatic whole-PDF OCR workflow. Recognition depends on the image models currently available.
Start translatingDrag files here
Loading available quota...
Online Text Translation
Rich text supported (bold, lists, headings). Translates automatically as you type, using the model and target language selected above.
Tokens used today: —Paid models: 1 credit per 1,000 input and output tokens combined; remaining tokens carry over. Free models use no credits.
When to use this workflow
Image-only scanned PDFs
If your reader cannot copy text from the PDF, export the pages you need as images. Working page by page helps identify missing lines, rotation or blur, and lets you start with the most relevant pages.
Photographed manuals and paper documents
Use a clear photograph with the page facing the right way. Keep the page flat, avoid reflections and obstructions, and crop unrelated background without cutting off text.
PDFs that already have OCR text
Selectable text can still contain recognition errors. Copy a passage and check for missing characters and incorrect reading order. If the text layer is usable, continue with ordinary PDF document translation.
How to translate and download your result
Check whether the PDF contains usable text
Try copying a complete sentence. If you can only select an image, or the copied text is garbled, address recognition first. If the text and reading order are correct, try Document translation.
Prepare a clear image of each required page
Use your PDF reader to export the page as PNG, JPEG or WebP; renaming the extension does not convert a PDF. Each image must be no larger than 20 MB. Split dense pages into sections if necessary and keep their reading order.
Upload an image to recognize and translate
Use Upload Document below or drop an image. Image settings appear automatically. Select available image and text translation models and set the target language. Choose Recognize and translate, or Recognize text only if you first need a transcription.
Verify recognized text before refining the translation
Check the image for missing lines, confused letters and digits, and the reading order of columns. Recognition errors can affect the translation, so review and correct both sides before saving through the available export options.
Limitations and checks before you use the result
- An image-only scanned PDF is different from a PDF with readable text. The document uploader does not guarantee whole-PDF OCR. Process exported page images, or use another OCR tool to create a searchable PDF first.
- Blur, skew, handwriting, overlapping stamps and dense tables can cause missing or incorrect text. A translation model cannot reliably recover details that are unreadable in the source image.
- Image translation needs available recognition and text translation models. Check the Image translation area for the current models, service availability and credit costs.
- You can adjust translated image text and its boxes, but this does not reconstruct the entire original document. Review each page separately; a single-image export does not preserve the layout of a complete scanned PDF.
Frequently asked questions
Can I upload an entire scanned PDF for OCR translation?
Do not treat an image-only scan as a normal text PDF. Export the required pages as images and recognize and translate them individually using Image translation. Alternatively, use another OCR tool to create a searchable PDF, check its text layer, and then upload it for document translation.
How can I tell whether my PDF needs OCR?
Copy a full passage from your PDF into a text editor. Missing text, garbled characters or incorrect reading order indicate that recognition or text-layer correction is needed. Some PDFs mix text pages with scanned pages, so check both types.
Which image formats are supported?
Image translation accepts PNG, JPEG and WebP images up to 20 MB each. Export the PDF page into one of those formats; changing its filename extension is not enough. Some recognition models require conversion or compression, which the page handles automatically where possible.
Does accurate OCR guarantee an accurate translation?
No. First check recognized text, numbers and reading order against the image, then review the translation for meaning and terminology. Compare easily confused characters such as 0 and O or 1 and l with the source, and recheck the corresponding translation after correcting recognition errors.
For allowance details or more help, visit the pricing pageand help and FAQ.
Start with your document type
Find the upload workflow, review options and checks that fit your file.
PDF translator
Translate papers, reports and manuals, review paired text and download a PDF.
View and get startedWord document translator
Upload DOCX or DOC and download an editable translation or bilingual Word file.
View and get startedScanned PDF and image translation
Learn how to export page images from scans for OCR, translation and review.
View and get started