OnDevice OCR Pro

Guide

After Text Recognition (OCR):
Name PDFs Automatically

Guides

Name and Sort Files

If you scan a lot of documents, you know that text recognition is only part of the work.

You can download OnDevice OCR Pro for free from the Mac App Store and test all features. The only limitation is a watermark in the output files.

The more time-consuming part is often the naming that follows. We show you how to name documents easily and according to your own requirements with OnDevice OCR Pro – from simple numbering, to automatically extracting values such as invoice numbers, case numbers, and property numbers, to naming based on complex terms such as senders, recipients, and company names using local AI.

And all of this while protecting your privacy. OnDevice OCR Pro works entirely locally, without transferring any data.

There are several ways to read values from the document

This is how you name PDFs automatically based on their content and file them directly in the right folder.

  1. Suggestions from the Document → OnDevice OCR Pro can analyze your documents and suggest values from them for the later file name. (See the picture on the right.)
  2. Value from Document → You can search the documents for custom terms, such as a case number or property number, and extract their values. If a term is found, the associated value (e.g. the individual property number) is read automatically and becomes part of the later file name.
  3. Search an Area in the Document → For the most difficult cases in complex layouts, you can have areas read by local AI models such as Apple Intelligence.
Example: the following terms with values were found in this invoice.

Frequently asked questions about automatic file naming

Can OnDevice OCR Pro name scanned PDFs automatically?

Yes, using values that you define yourself and that appear in the documents, such as the date, invoice number, case number, property number, etc. For complex file names with the recipient, subject, or sender, you can also integrate (local) AI models such as Apple Intelligence, Gemma 4, or Qwen into the app via the Ollama interface.

Which values can be used for the file name?

Recognized dates, numbers following a term (invoice number, customer number, property number), your own search terms, fixed text, a sequential number, and – with Apple Intelligence and other local AI models – also the sender, recipient, company name, etc.

Does this also work with different invoice layouts?

Yes, by using search terms instead of fixed positions. For difficult layouts, you can have an area of the document read by local AI. A separate, individual template can be created for each type of document. Rules can be used to select the matching template immediately based on the recognized text.

What happens if a value is not found in the document?

The file is still processed. The missing component is left out of the file name; the other components remain. If nothing is left, the file keeps its original name. The app shows a notice indicating which value could not be read.

Are my documents sent to a cloud AI for this?

No, everything runs locally on your Mac, including Apple Intelligence and Ollama. When using Ollama, however, it is your own responsibility to select only local models for naming.

Can I also have the documents sorted into suitable folders at the same time?

Yes, with rules, the documents can be saved in folders you specify depending on the recognized content.

General questions about pricing, licensing, languages, and system requirements are answered in the FAQ on the home page.

Name your documents automatically – with OnDevice OCR Pro.

Download on the Mac App Store

Compatible with macOS 14.6 and later.