- JSON
- CSV
- TEXT
- HUMAN
Short answer
How can I get data out of a scanned PDF?
Your AI uses OCR to recognize text within images and then converts that text into JSON or CSV files. You'll get structured data ready for your spreadsheets or databases.
Where your data lands.
The final outputs.
The AI processes the file and delivers it in the format you need.
JSON
Structured JSON files
The AI converts document fields into a JSON object. This is ready to be used in other software or APIs.
CSV
Spreadsheet-ready CSVs
Scanned tables become CSV files. You can open these immediately in any spreadsheet app for review.
TEXT
Raw text strings
The AI pulls all readable text from the document. This works well if you just need the content for a summary.
HUMAN
Manual verification
You decide if the extracted data needs a second look. The AI does the extraction, but you keep the final say.
The workflow
What your AI does when a scanned PDF arrives.
The AI follows a sequence to turn pixels into organized information.
Read the image
The AI looks at the scanned document to identify characters and layout.
ocr_imageExtract text
The agent pulls the identified characters into a plain text format.
pdf_to_textStructure the data
The AI organizes the text into a specific JSON schema.
pdf_to_jsonFormat for tables
The agent converts the document content into a CSV file.
pdf_to_csv
Try it
Copy these to start.
Use these prompts to trigger the extraction process.
Starting points
Replace the bracketed text with your specific file names or target formats.
PDF.co Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_uVxMMI3j4I5XvY70ujxxyTdKG0jtWoShDwpFOIBv/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — PDF.co capabilities are ready to use.
{
"mcpServers": {
"pdfco-mcp": {
"url": "https://edge.vinkius.com/vk_preview_uVxMMI3j4I5XvY70ujxxyTdKG0jtWoShDwpFOIBv/mcp"
}
}
}Use PDF.co to run OCR on the invoice.pdf file and give me the results as JSON.
Convert the scanned receipt in my documents folder to a CSV file.
Extract all the text from the scanned contract.pdf so I can summarize it.
Turn this scanned table into a JSON object for my database.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect PDF.co once, then ask.
You only need to link your account once. Your credentials stay encrypted, and you can immediately start asking your AI to process files.
Connect PDF.co to your AIFAQ
How this task behaves.
- 01
Can the AI fix blurry scans?
No. The AI relies on the clarity of the original image to perform OCR. If the scan is too blurry to read, the extraction will fail.
- 02
Will the AI change the original PDF file?
No. The tools only read the existing file and create new JSON, CSV, or text outputs. Your original document remains untouched.
- 03
Can I convert a PDF to an Excel file directly?
The AI converts files to CSV. You can open any CSV in Excel without issues.
- 04
Does the AI handle handwriting?
The accuracy depends on the scan quality, but the OCR capability is designed to recognize text within image-based documents.
- 05
Can the AI delete files after processing?
No. This MCP only has tools for reading and converting data. It cannot delete or modify files on your system.
More questions about PDF.co? The Connector page answers them. See everything the PDF.co Connector can do
Connect PDF.co to Claude, Cursor, ChatGPT & more
