PDF to JSON
Extract structured PDF text, metadata, page dimensions, and coordinate blocks into clean JSON.
Extracts document hierarchy, text runs, font definitions, and metadata into JSON structured data entirely within browser memory.
The FileTools PDF to JSON utility extracts digital text tokens, font metadata, typographical scaling, and exact spatial bounding coordinates from PDF documents into structured JSON format. Ideal for developers, data scientists, and automation engineers building OCR verification pipelines, document parsing workflows, and AI extraction pipelines completely client-side.
Key Challenges Solved
- ✓ Structured invoice tables, line items, and financial figures locked inside unstructured PDF streams.
- ✓ Manual data entry from PDFs into ERP systems being slow, labor-intensive, and prone to typing errors.
- ✓ Needing clean structured JSON schemas for programmatic ingestion into databases and modern APIs.
Who Is PDF to JSON Built For?
Software engineers and DevOps teams piping PDF invoice data directly into billing databases
Fintech developers building automated accounting feeds from bank statements and tax documents
Data analysts extracting survey responses and tabular data from published research PDF reports
Spatial PDF Tokenization: Converting Fixed Typography into Hierarchical JSON
How 2D affine transformation matrices are mapped into structured machine-readable JSON schemas
What We Verified
- ✓ Affine matrix decoding: Reads the 6-element transformation array [a, b, c, d, e, f] to extract absolute horizontal and vertical baseline coordinates for every word glyph.
- ✓ Structured RFC 8259 output: Emits hierarchical page arrays containing width, height, line bounding boxes, and nested text tokens with calculated font points.
- ✓ Developer-ready ingestion: Produces clean JSON objects designed for automated ETL pipelines, local LLM context ingestion, and database indexing.
- ✓ Zero-cloud privacy guarantee: Extraction runs entirely inside Mozilla PDF.js in an isolated Web Worker; sensitive corporate financial records are never sent to external AI APIs.
Why This Happens (Technical Architecture)
A PDF document does not store paragraphs or semantic HTML blocks; it stores discrete string rendering instructions associated with font objects and coordinate translation matrices. FileTools uses pdfjs-dist getTextContent() to walk page display lists, groups adjacent glyphs sharing identical Y-baselines into line tokens, calculates approximate font heights in CSS points, and normalizes coordinates relative to the page viewport. The output is formatted into standardized JSON with configurable indentation.
Key Features & Benefits
Spatial Coordinate Mapping
Extracts exact X, Y coordinates, width, and height for every text block and line.
Typographical Metadata
Captures font family names, font sizes, transformation matrices, and text directionality.
Page-by-Page Hierarchy
Structures data hierarchically: Document → Pages → Lines → Word Tokens.
Zero API Keys & Total Privacy
Extract financial statements and proprietary invoices locally without third-party API exposure.
How to Use PDF to JSON
- Upload your PDF document into the PDF to JSON dropzone.
- Configure extraction options: include bounding coordinates, font styles, or line grouping.
- Inspect the live formatted JSON tree or raw JSON editor.
- Copy the JSON to your clipboard or download a structured .json file.
Common Use Cases
Invoice & Receipt Extraction Pipelines
Extract structured line-item positions and amounts from commercial PDF invoices into structured JSON.
AI & LLM Document Ingestion
Provide large language models with spatial bounding box context for complex multi-column documents.
Form Field Coordinate Discovery
Identify exact PDF coordinates to programmatically populate AcroForm fields in automated pipelines.
Continue Your Workflow
Recommended logical next steps after using PDF to JSON:
JSON Formatter & Validator
Format, validate, minify, inspect, and query JSON with interactive tree views and zero server uploads.
Use JSON Formatter & Validator →CSV to JSON Converter
Convert CSV and TSV spreadsheets into structured JSON arrays and objects with type auto-detection.
Use CSV to JSON Converter →PDF to Excel
Convert PDF tables and financial data into real Microsoft Excel (.xlsx) spreadsheets.
Use PDF to Excel →You May Need This Before
Common preparation and prerequisite steps before PDF to JSON:
OCR PDF to Searchable
Convert scanned and image-only PDFs into searchable, selectable documents using in-browser OCR.
← Prerequisite: OCR PDF to SearchablePDF to Excel
Convert PDF tables and financial data into real Microsoft Excel (.xlsx) spreadsheets.
← Prerequisite: PDF to ExcelPDF to Text
Extract clean selectable text from PDF documents with instant copy and TXT file download.
← Prerequisite: PDF to Textpdfjs-dist TextContent API Parsing
FileTools utilizes pdfjs-dist getTextContent() stream parsing. Each text item is evaluated for its transformation matrix (transform array) to determine absolute canvas coordinates, baseline font size, and rotation angle relative to the page viewport.
Extracting Invoice Total and Line Items
A developer needs to parse 50 PDF invoices into JSON data for accounting software without sending invoices to cloud OCR APIs.
The developer uploads the invoice, downloads structured JSON in 1 second, and imports it directly into their database.
Important Operational Notes & Realistic Limitations
- Scanned image-only PDFs must be converted with OCR PDF first to generate machine-readable text layers.
- Vector graphical shapes and raster images are omitted; only text streams and typographical bounds are serialized.
- Multi-column layouts require downstream coordinate grouping by X/Y coordinates to establish linear reading order.
Frequently Asked Questions
Can I convert PDF content to JSON?
Yes. The pdf structured data extractor turns pages into clean blocks, so you can convert PDF to JSON and extract PDF data as JSON for scripts, databases, or automation pipelines.
What does the extracted PDF JSON output structure look like?
The output follows a valid RFC 8259 JSON hierarchy containing an array of pages (with pageNumber, width, height) and nested textBlocks (with text, x, y, width, height, fontSize, and fontFamily). You can inspect and validate the output using our JSON Formatter (/json-formatter).
Can I use this JSON for automated NLP, indexing, or web scraping pipelines?
Yes. The structured JSON format is ideal for developer pipelines, automated indexing, AI prompt context, and database ingestion.
Does the JSON output include font metadata like bold, italic, and font size?
Yes. The parser captures font family tokens, approximate font sizes in points, and typographic weights alongside positional bounding boxes.
What coordinate system is used for bounding boxes?
Coordinates use standard PDF points (1/72 inch). The origin [0, 0] starts at the bottom-left of the page viewport, and transformation matrices map glyph offsets to page width and height.
Can this JSON be used for AI RAG pipelines or local LLMs?
Yes. By retaining spatial coordinates and line breaks, extracted JSON preserves semantic document structure, making it ideal for feeding structured context to local LLMs, vector embeddings, or automated invoice parsing models.