Tiếng Việt

Image to text (OCR)

Recognise the text in photos of documents, screenshots or scanned PDFs so you can edit and copy it. Vietnamese accents are supported.

PDF (up to 30 pages) or JPG, PNG, WEBP images; pick several images to read several pages in order.

The first time, your browser downloads the Vietnamese recognition data (about 5 MB); later runs reuse it.

Files are processed on your device and never uploaded.

How to use it

  1. Choose JPG, PNG or WEBP images or a scanned PDF; several images make several pages.
  2. Wait while each page is recognised; progress is shown.
  3. Check and edit the text in the result box.
  4. Copy it, or download a Word (.docx) or TXT file.

Getting accurate results

Image quality matters most:

  • Shoot straight on, with the page filling the frame.
  • Use even light without shadows or glare.
  • Scan at 200–300 DPI.
  • Make sure the image is upright.

Limits

Printed text works well. Handwriting, text under stamps, tiny or faint print may be misread, and tables come out as lines of text. Check names, numbers and dates.

Frequently asked questions

How long does a page take?

Usually 5 to 20 seconds, depending on your device and the amount of text.

Is my file uploaded?

No. Recognition runs in your browser; the first run downloads about 5 MB of recognition data.