OCR — searchable PDF

Make a scanned PDF searchable: pages without a text layer are OCR'd with Tesseract and gain an invisible text layer; pages that already have text are passed through and reported as skipped.

Free on the API and MCP server — generous limits, files deleted within an hour, content never logged.

Server

This one runs on our servers — under the same privacy rules.
Files are virus-scanned before processing, stored encrypted, and deleted within 1 hour. Content is never logged. And it's free: 250 operations a day, every day, plus burst on spare capacity.

OCR — searchable PDF needs server-grade machinery (heavy conversion, OCR or cryptography the browser can't do safely), so it runs through the REST API and the MCP server instead of an uploader. One free key drives both — get yours in about a minute.

REST API

curl -X POST https://api.pdfmcp.org/v1/jobs \
  -H "X-API-Key: pdfmcp_live_YOUR_KEY" \
  -F [email protected] -F op=ocr_pdf -F language=eng -F dpi=300

Submit → poll → download. Full walkthrough in the REST API quickstart guide.

MCP (for AI agents)

{
  "mcpServers": {
    "pdfmcp": {
      "url": "https://mcp.pdfmcp.org/mcp",
      "headers": {
        "X-API-Key": "pdfmcp_live_YOUR_KEY"
      }
    }
  }
}

Paste into Claude, Cursor, VS Code, ChatGPT and 9 more clients — setup guides for every major MCP client.

By using the API or MCP server you agree to the Terms of Service and Acceptable Use Policy. Files you send are deleted within an hour; see the Privacy Policy.

How ocr — searchable pdf works here

Make a scanned PDF searchable: pages without a text layer are OCR'd with Tesseract and gain an invisible text layer; pages that already have text are passed through and reported as skipped. Server-side, the job passes the security inspection gauntlet first (antivirus, structural and active-content checks), runs in a sandboxed worker, and the result waits for you for exactly one hour.

People often pair this with pdf to text and pdf info & metadata — same deal there: free, private, no sign-up.

Parameters

Parameter	Type		Default	Description
`language`	string	optional	eng	Tesseract language(s), e.g. 'eng' or 'eng+deu'. Must be installed on the server.
`dpi`	integer	optional	300
`pages`	string	optional	—	1-based page selection, e.g. '1-3,5,8'. Default: all pages.

FAQ

Is OCR — searchable PDF free?

Yes — completely free, with no watermark, no sign-up and no trial clock. On the API/MCP side every key gets 250 guaranteed operations a day plus unlimited burst operations on spare capacity. See the pricing page for the full (short) policy.

What happens to my file?

You send it to the pdfmcp API/MCP server over HTTPS. It is virus-scanned and inspected before processing, stored in an encrypted temporary volume, and automatically deleted within 1 hour of the job finishing. We never log file content or file names.

Can I automate OCR — searchable PDF with an AI agent or code?

Yes — the exact same operation is a tool on the pdfmcp MCP server (for Claude, Cursor, ChatGPT and other agents) and an endpoint on the REST API. One free key works for both; the guides walk you through setup in about a minute.

Are there limits?

Files up to 100 MB and 500 pages. Web tools are additionally bounded by your device's memory. The caps exist to stop abuse and keep the shared machine fair — they are not a paywall, and there is no paid tier to unlock.

Browse the extract & ocr hub or the full tool index. Building something bigger? Read the features tour and the free usage policy.

OCR — searchable PDF

REST API

MCP (for AI agents)

How ocr — searchable pdf works here

Parameters

FAQ

Related tools