Stella
PDFs and Excel

Make a scanned PDF searchable

A 6-page scanned lease you can search, looking exactly like the scan.

Download StellaFree with the plan you already have.

A real run on 13 November 2026, on a Linux computer, using Claude Code with Opus, on a Claude plan. The waiting is sped up; it took 1 min 41 s in real time.

The short answer.

A scan is a picture of text, so search finds nothing until OCR (text recognition) adds an invisible text layer. Free routes: OCRmyPDF from the command line on any system, or PDF24 Creator's OCR tool on Windows; Acrobat's Scan & OCR does it if you pay. Preview on a Mac can't. Stella (stella.sh) runs OCR on your computer, saves a searchable copy next to the scan and checks that searches work.

Stella does this on Windows, Mac and Linux.

What happened.

A searchable copy of the 6-page lease in about 1 minute 40 seconds. Checked afterwards against an answer key written before the run: every page has a text layer, all 15 key phrases can be found by search, the pages look identical to the scan, the page count is the same and the file (2.6 MB) is no more than three times the original's size.

The lease is fictitious, scanned for this run as greyscale page images with slight skew, blur and noise. The prompt says Ctrl+F because it ran on Linux. The work is sped up 4×.

How Stella does it.

  1. Finds the scanned lease in Downloads: 6 pages of page images with no text in them
  2. Installs OCRmyPDF, a free text-recognition tool, and runs it on a copy, not the original
  3. Keeps each page looking exactly like the scan, with the recognised text laid invisibly underneath so you can search and select it
  4. Checks the result by searching for the names, the address and the dates
  5. Warns you about a catch it found: the rent is written only in words, so searching “14,200” won't find it

The prompt.

The scanned lease in my Downloads isn't searchable. I can't find anything in it with Ctrl+F. Can you make it searchable?

Paste it into Stella and press Enter.

This is the exact prompt from the recording. If your files are somewhere else, say where.

Do it yourself.

  1. Check whether it really is a scan

    Open the PDF and try to select a line of text. If you can't, or the whole page highlights as one block, it's a picture. A PDF you can select but still can't search may use an odd font encoding; OCR fixes that too.

  2. Any system: OCRmyPDF

    The free OCRmyPDF adds a text layer and keeps the page images: ocrmypdf --language eng scan.pdf searchable.pdf. Add --deskew for crooked pages. On a Mac it usually comes from Homebrew (brew install ocrmypdf), on Linux from your package manager.

  3. Windows: PDF24 Creator

    Install the free PDF24 Creator, open its OCR tool, add the PDF, choose the document's language and start. It saves a new searchable PDF and runs on your computer, so the scan isn't uploaded.

  4. Mac without Terminal

    Preview can't add a text layer. A free PDF app with OCR, such as PDFgear (its OCR to Searchable PDF option), does it in a few clicks. Acrobat (paid) has Scan & OCR, Recognize text, In this file.

  5. Test the result

    Search for a name, a date and a number from different pages. OCR misreads small print and smudged figures (1 and 7, 0 and O), so check any amount that matters against the page itself.

Where it stops.

  • OCR can misread small print, handwriting and smudged numbers. The test scan was clean enough that every search worked, but check figures that matter on your own documents.
  • Handwritten notes and signatures stay as pictures; they aren't turned into text.
  • Recorded on Linux, where the Tesseract text-recognition engine was already installed and Stella added OCRmyPDF. A Mac usually has neither, and installing them normally needs Homebrew, so Stella would take a different route there, such as macOS's built-in text recognition; that wasn't recorded.
  • Recognition runs on your computer, but the recognised text goes to the AI model you use in Stella when it checks the result. For confidential papers, choose your model accordingly.

Questions.

Why can't I search my scanned PDF?

A scanner or phone app saves each page as a photo. There's no text in the file for search to find until OCR adds a hidden text layer behind the picture.

Can Preview make a PDF searchable?

No. Preview can search and copy text that's already in a PDF, but it can't recognise the text in a scan and save it into the file. Use one of the tools above.

Is there a free way to OCR a PDF without uploading it?

Yes: OCRmyPDF on any system, PDF24 Creator on Windows, or Stella with the plan you already have. Online OCR sites work, but a lease or medical letter is exactly what you'd rather keep on your computer.

Will OCR make the file bigger?

A little: the text layer is small next to the page images. In the recording the copy stayed within three times the original. If it's too big to send, compress the PDF; to pull figures into a sheet, see bank statement PDF to Excel.

Is Stella free for this?

Yes. Free with the ChatGPT or Claude plan you already have. Sign in with your ChatGPT plan, or use your Claude plan through Claude Code, and this task costs nothing extra. New accounts also get a small free trial of Stella’s own models.

Can I control what Stella changes on my computer?

Yes. Stella works on Windows, Mac and Linux with your own files and shows each step in the chat, so you can watch it and stop it at any point. Tell it what it may not do, as the prompts on these pages do (“don't delete anything without asking”). In the Downloads recording it deleted nothing until it was told to go ahead. It also refuses commands that would wipe a whole disk, your home folder or system folders.

Checked against these sources on 13 November 2026.

Stella