Searchable PDF OCR

Recognize English, Spanish, French or German PDF scans, review words and export searchable PDF plus UTF-8 text.

On-device No sign-up Free to use

Export current result controls

  1. 1 Add files
  2. 2 Arrange/Options
  3. 3 Download

Drop your PDF here

or choose one from your device

Up to 20 MiB and 100 source pages. Recognize at most five selected pages per run. Review words and positions before export.

    ?

    Language models load from this site. Choose the language used in the scan.

    ?

    Optional, such as 2-4, 8. Select at most five pages. Leave empty only when the PDF has five pages or fewer.

    Advanced

    Drop your file here

    or choose one from your device

    A TTF or OTF font up to 8 MiB. The default Noto Sans font covers Latin, Greek and Cyrillic. Use a font that contains every character in your text.

      Result

      Showing an example.

      Add reviewed word positions to a simple English scan.

      A one-page scan reading WAVE FOUR produces a rasterized PDF with two positioned selectable words after review.

      Loading the tool. Your result will appear here.

      https://usefreetools.net/searchable-pdf-ocr · 20 November 2026

      Was this helpful?

      How to use Searchable PDF OCR

      1. Choose an unencrypted PDF up to 20 MiB and 100 source pages. Select up to five pages and the scan language.
      2. Set any quarter-turn rotation, recognize scans, then review every word and its pixel box against the page images.
      3. Correct, remove or add words. The editor preserves Unicode. Wait for preview updates and confirm review on all selected pages.
      4. Export the searchable PDF and UTF-8 text file. Reopen the PDF to verify search, selection order and every correction.

      Example: Searchable PDF OCR

      Add reviewed word positions to a simple English scan.

      You add
      Rotate scans before OCR: 0 I reviewed the recognized words and their positions on every page: true PDF file: One scanned page reading WAVE FOUR; recognize and review its word boxes before confirming.
      You get
      A one-page scan reading WAVE FOUR produces a rasterized PDF with two positioned selectable words after review.

      Supported inputs and limits

      • Bundled English, Spanish, French and German language models load from this site.
      • One PDF up to 20 MiB and 100 source pages; at most five selected pages per run.
      • Each page is capped at 12 million raster pixels and 3,000 reviewed words.
      • Each word allows 1–100 printable Unicode characters; boxes must fit the page image.
      • The searchable PDF is rasterized at 144 dpi with an invisible text layer and capped at 50 MiB.
      More details

      Standard PDF fonts handle supported Latin text; the bundled Noto Sans fallback covers Latin, Greek and Cyrillic.

      A custom TTF or OTF up to 8 MiB can extend coverage, and missing glyphs are rejected.

      The text download preserves reviewed Unicode.

      OCR can omit, misread or misplace words.

      Forms, links, signatures, vectors and source metadata are not copied, and an invisible text layer does not establish PDF accessibility.

      Where your input is processed

      This tool does not upload your inputs. See the limits for anything saved on your device.

      Questions about Searchable PDF OCR

      Which scan languages work?

      Choose English, Spanish, French or German. Language data loads from the site, and recognition runs in the browser.

      Can I correct non-English words?

      Yes. Reviewed words and the UTF-8 text download preserve Unicode. PDF export also needs a font containing every character and reports missing glyphs.

      Are original vectors preserved?

      No. Visible selected pages are rendered images at 144 dpi. Review the saved PDF and the separate text file.

      Page updated · Project managed by Tony Hines · Report a problem with Searchable PDF OCR