Preview: this language version was translated with machine assistance and has not been reviewed by a native speaker. Search engines are asked not to index it.

How documents are processed on your computer

Short answer: EnjoyPDF handles every step of document processing on your computer. The only feature that sends document text out is the optional cloud AI translation, with a provider you choose and confirm. Recognition cache, project files and settings are stored locally; this page lists where they are and how to clear them.

Last updated: Checked against version: EnjoyPDF 2.0

Data flow

  1. Read: PDFs or images are read from your local disk. A password is used only for the current operation in memory; it is not stored in settings or logs.
  2. Recognize: Built-in or installed recognition models run on your local CPU. No online service is called.
  3. Replace or redact: This happens in memory. The original file is not overwritten.
  4. Save and verify: The output is written to the location you choose, then reopened locally for read-back verification or residue checks.
  5. Record: An operation record (manifest.json) is written next to the output file. It does not contain the replaced or removed text.

What is stored on your computer

ItemLocationContentsHow to clear
Recognition cache%LOCALAPPDATA%\PDFTextReplacer\ocr-cache\Recognition results, stored by file content hashDelete the folder
Project files%LOCALAPPDATA%\PDFTextReplacer\projects\*.pdtwKeywords, settings and selection state, used for crash recoveryDelete the file
Settings%LOCALAPPDATA%\PDFTextReplacer\settings.jsonInterface and recognition options; no file content, keywords or passwordsDelete the file
AI platform keys%LOCALAPPDATA%\PDFTextReplacer\keys.jsonEncrypted with the current Windows account (DPAPI)Delete in “Manage platforms…”
Language packs%LOCALAPPDATA%\PDFTextReplacer\modelsRecognition model filesDelete the file

The recognition cache becomes invalid automatically when file content changes or a model is upgraded.

License

The current development build does not include licensing or activation yet. The design goal is activation and checks on your computer, without going online; this section will be updated when the release version ships.

When the software goes online

The current version has only two kinds of network activity: downloading a language pack when you click, and cloud AI translation that you choose to use (which sends the text to be translated to the provider you chose). Update checks and error reports are planned and not implemented yet. See Network behavior, and check it yourself using Verify that the software does not upload documents.

Try it first with your own file

After downloading, open a scan, enter a keyword and click "Find". Finding, replacing and saving need no network connection.