Extract structured JSON from any PDF — metadata, page content, and text lines with coordinates. Instantly, for free, right in your browser. Nothing is uploaded.
Drop your PDF here or click to browse
100% client-side. Your file never leaves your browser.
Your PDF is processed entirely in your browser. Nothing is uploaded to any server.
Get clean JSON with metadata, page dimensions, text content, and individual lines with coordinates.
Runs on any modern browser — desktop, tablet, or mobile. No software to install.
JSON is the language of pipelines. Once a PDF becomes JSON, it stops being a locked document and becomes data you can query, filter, validate, store in a database, or pass to the next step of a workflow — no PDF library required downstream.
Unlike plain text, this output keeps the geometry of the document: every line comes with x/y coordinates, font size, and font name. That is what you need for layout analysis — finding the value next to a label on an invoice, detecting headers from font sizes, splitting columns, or building training datasets for document AI models.
{
"metadata": {
"title": "Q2 Financial Report",
"author": "Finance Team",
"creationDate": "2026-04-02T09:14:00Z"
},
"pageCount": 12,
"pages": [
{
"pageNumber": 1,
"width": 595.28,
"height": 841.89,
"text": "Q2 Financial Report...",
"lines": [
{
"text": "Q2 Financial Report",
"x": 56.7, "y": 780.2,
"fontSize": 24, "fontName": "Helvetica-Bold"
}
]
}
]
}Drag and drop or click to select a PDF file from your device. It is read locally — never uploaded.
The engine reads every page and structures metadata, text, and line coordinates into clean, machine-readable JSON.
Copy the JSON to your clipboard or download it as a .json file, ready for your scripts and pipelines.
The output includes document metadata (title, author, creation and modification dates), and for each page: its dimensions, the full text content, and every text line with its position coordinates and font information.
No. The entire extraction happens in your browser using JavaScript (PDF.js). Your file never leaves your device — safe for invoices, contracts, and any confidential document.
Coordinates let you reason about layout programmatically: detect columns, find a value positioned to the right of a label (like a total on an invoice), rebuild reading order, or crop a region of interest. They are the raw material of document layout analysis.
Yes. Every conversion produces the same shape — metadata plus an array of pages, each with dimensions, text, and lines — so you can write parsing code once and run it on any PDF.
No — scanned PDFs are images without a text layer. They need OCR, which is available through the ParseDocu API for both scans and photos of documents.
Not directly: you get every cell as a positioned text line, which lets you rebuild simple tables from the coordinates. For ready-to-use rows and columns, the ParseDocu API has a dedicated table extraction endpoint.
Yes — free, unlimited, no account required. It is our way of showing what ParseDocu does before you try the API.
For automated pipelines, use the ParseDocu API: one POST request returns structured output at scale, with OCR and table extraction included, and connectors for Zapier, Make, and n8n. The free tier includes 1,000 credits.
The ParseDocu API turns digital and scanned PDFs into structured data at scale — via REST, Zapier, Make, or n8n. Start with 1,000 free credits.
Get 1,000 free API creditsNo credit card required · Read the API docs
More free document parsing tools — all private, all in your browser.
