• SEARCHABLE
  • ANALYZABLE
  • EDITABLE
  • HUMAN

How to make scanned documents searchable with your AI

Connect PSPDFKit (Nutrient) to your AI client to turn static images into data. Once connected, your agent handles the text recognition for you.

Ask AI about this page

Short answer

How can I make my scanned PDFs searchable using AI?

Your AI uses the ocr_document tool to recognize characters in your images and create a text layer. This lets you search, highlight, and copy text from files that used to be just flat pictures. You'll see how the process works below.

The results.

What your files become.

Turning images into text changes how your agent interacts with your data.

SEARCHABLE

Find specific text

The AI identifies characters within your images. You can then search for specific keywords or phrases across your entire document library.

ANALYZABLE

Query document content

Once the text is extracted, your agent can summarize long scanned reports or pull specific data points from old invoices.

EDITABLE

Copy and paste text

The tool creates a text layer over the image. This lets you highlight and copy text directly from a scan into your chat.

HUMAN

Verify accuracy

You stay in control of the final output. You'll need to check the AI's work for accuracy on messy or handwritten scans.

The workflow

What your AI does when a scan arrives.

The AI moves from raw pixels to structured text through a single specialized operation.

  1. Visual analysis

    The agent looks at the uploaded file to identify text regions and character shapes.

    ocr_document

  2. Character conversion

    The tool converts those visual shapes into digital characters that your AI can read.

    ocr_document

  3. Layering text

    A searchable text layer is mapped onto the original image so the visual layout stays the same.

    ocr_document

  4. Data readiness

    The document is now ready for your agent to perform searches or summaries.

    ocr_document

Try it

Copy these to start.

Use these prompts to trigger the OCR process in your chat.

Starting points

These are ways to kick off the process. Swap out the bracketed text for your actual file names or folder paths.

PSPDFKit (Nutrient) Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_GJ5munaGM52GsFLuBPidwWhHisnl0Jr4DQUgzLCy/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — PSPDFKit (Nutrient) capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "pspdfkit-nutrient-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_GJ5munaGM52GsFLuBPidwWhHisnl0Jr4DQUgzLCy/mcp"
    }
  }
}
Copy into chat04

  • Run OCR on the scanned invoice in my Documents folder.

  • Make this PDF image searchable so I can find the contract number.

  • Extract the text from the scanned receipt I just uploaded.

  • Search through the scanned project notes for any mention of 'budget'.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect PSPDFKit (Nutrient) once, then ask.

Just link your account once. Your credentials stay encrypted, and you can immediately start asking your agent to process files.

Connect PSPDFKit (Nutrient) to your AI

FAQ

How the OCR task behaves.

  • 01

    Can the AI read handwriting?

    The accuracy depends on the clarity of the scan. While it can often read handwriting, you should always verify the results manually.

  • 02

    Does this change my original image file?

    No. The tool adds a text layer to make the file searchable, but it doesn't alter the original visual pixels of your document.

  • 03

    Can the AI delete my scanned documents?

    No. The available tools only allow for reading and processing files through OCR; there is no tool to delete or modify your files.

  • 04

    Will it work on low-quality photos?

    It works best on clean, high-resolution scans. If a photo is too blurry or dark, the text recognition might fail.

  • 05

    Can I edit the text directly in the chat?

    No. The AI can read and summarize the text, but you have to use a PDF editor if you want to change the actual content of the file.

  • More questions about PSPDFKit (Nutrient)? The Connector page answers them. See everything the PSPDFKit (Nutrient) Connector can do

Connect PSPDFKit (Nutrient) to Claude, Cursor, ChatGPT & more