Fidument Tools · Beta version

Image to text — OCR

Recognize printed Polish, English or German text.

Text in a photo or scan cannot be selected like ordinary text.

Who is it for?

  • People preparing documents and work materials

Input

  • Up to 16 MiB, 20 MP, 8192 px. JPG/PNG/WebP, no animation or handwriting. First use downloads a language model. Check the result against the source.

Output

  • TXT UTF-8

Capabilities

  • Choose a clear printed-text image and the matching PL, EN or DE language.
  • Rotate by 90°, 180° or 270° if needed, then recognize. First use downloads the language model.
  • Compare against the source, especially numbers and names. Copy text or download UTF-8 TXT.

Limitations

  • Up to 16 MiB, 20 MP, 8192 px. JPG/PNG/WebP, no animation or handwriting. First use downloads a language model. Check the result against the source.
  • Files are processed on your device. Static libraries and models are downloaded. Saving finished files to private account history requires sign-in and explicit confirmation.

Example use

An own “Fidument / Zamówienie 123 / Blacha 40 × 60 mm” sample checks names, numbers and Unicode; results always need review.

Questions and answers

Where is my file processed?

Files are processed on your device. Static libraries and models are downloaded. Saving finished files to private account history requires sign-in and explicit confirmation.

Is the tool published yet?

Yes. The tool is available in beta. Check its limitations and use the launch button on this page.

Related tools

PDF ↔ Images

Convert PDF pages to PNG or JPG and combine PNG or JPG files into a PDF in your browser.

Resize and compress images

Resize images, adjust JPEG or WebP quality and convert files in your browser. Compare the actual input and output file sizes.

Image to text — OCR

Check access and export conditions in the app. A product description does not grant access.