# opendataloader-project/opendataloader-pdf

PDF Parser for AI-ready data. Automate PDF accessibility. Open-source.

Repository: https://github.com/opendataloader-project/opendataloader-pdf
Canonical: https://ross.abutalabs.com/products/opendataloader-pdf
Homepage: https://opendataloader.org
Language: Java
License: Apache-2.0
License Family: permissive
Topics: json, markdown, pdf, ai, document-parsing, html, pdf-converter, tables, pdf-parser, rag, accessibility, bounding-box, eaa, ocr, pdf-accessibility, pdf-extraction, pdf-ua, tagged-pdf, ocr-recognition, a11y
Last push: 2026-08-26T13:05:00+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 99, longevity 34
- inputs: {"age_days": 477, "days_push": 7, "days_rel": 8, "gap_med": 1, "n_releases_24m": 54}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 28817, forks 2750 (observed 2026-08-28T04:11:52.872269+00:00)

## What it is
OpenDataLoader PDF is an open-source (Apache-2.0) PDF parser that converts PDFs into AI-ready Markdown, JSON with per-element bounding boxes, and HTML, featuring XY-Cut++ reading order, table extraction, and optional hybrid-mode OCR for scanned documents. It also automates PDF accessibility by auto-tagging untagged PDFs into Tagged PDF as a foundation for PDF/UA compliance, and ships local AI-safety filters that strip hidden prompt-injection text before documents reach LLM agents.

## Use cases
- convert PDF documents to markdown for a RAG pipeline
- extract tables from PDFs while preserving row and column structure
- parse multi-column or academic PDFs with correct reading order
- get bounding box coordinates for every PDF element to support citations
- run OCR on scanned PDFs locally without cloud APIs
- automate PDF accessibility tagging for EAA or Section 508 compliance
- detect hidden prompt injection text in PDFs before feeding them to agents

## When to choose
- You need accurate, deterministic PDF-to-Markdown/JSON conversion for RAG or LLM pipelines with correct reading order, tables, and source coordinates
- You must process documents 100% locally and privately on CPU, with no data transmission and no GPU required
- You handle scanned or complex pages and want optional AI/OCR hybrid enhancement while keeping simple pages fast
- You need to scale PDF accessibility remediation by auto-tagging untagged PDFs toward PDF/UA and validating with veraPDF

## When to avoid
- You only need to view, annotate, or fill PDF forms rather than extract structured data
- Your environment cannot run Java 11+, since the core parser is JVM-based
- You need PDF/UA export or the accessibility studio, which are gated behind the enterprise add-on rather than the open-source core
- You need a browser-side or pure-JavaScript in-process parser with no external runtime or backend server

## Facets
- artifact type: library
- maturity: active
- function: parser, pdf, ocr, rag
- domain: pdf, accessibility, developer-tools, files
- platform: cross-platform, cli, python, jvm
- tags: pdf-parser, document-parsing, rag-pipeline, pdf-accessibility, tagged-pdf, pdf-ua, eaa-compliance, bounding-boxes, reading-order, table-extraction, markdown-conversion, llm-ready-data, prompt-injection-defense, layout-analysis, a11y, retrieval-augmented-generation, nodejs

## Member repositories
- opendataloader-project/opendataloader-pdf (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:52.872269+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:54:19.141207+00:00, confidence not recorded.
  - readme: https://github.com/opendataloader-project/opendataloader-pdf (fetched 2026-08-28T04:11:52.872269+00:00, sha 2304f60bba1a)
  - homepage: https://opendataloader.org (fetched 2026-08-29T07:50:39.575049+00:00, sha 23c88043f31e)
  - site_page: https://opendataloader.org/docs (fetched 2026-08-29T07:50:39.584446+00:00, sha 3a1541cedc18)
  - site_page: https://opendataloader.org/docs/quick-start-python (fetched 2026-08-29T07:50:39.586759+00:00, sha 668c8421d2aa)
  - site_page: https://opendataloader.org/docs/reading-order (fetched 2026-08-29T07:50:39.589629+00:00, sha 0e579ae96cb6)
  - site_page: https://opendataloader.org/docs/hybrid-mode (fetched 2026-08-29T07:50:39.592026+00:00, sha d23722d5ec13)
  - site_page: https://opendataloader.org/docs/reference/json-schema (fetched 2026-08-29T07:50:39.594750+00:00, sha de2eece53ed9)
  - site_page: https://opendataloader.org/docs/ai-safety (fetched 2026-08-29T07:50:39.598106+00:00, sha 4016a8b17e2a)
  - site_page: https://opendataloader.org/docs/benchmark (fetched 2026-08-29T07:50:39.601672+00:00, sha d8bd0ae2cf6c)
  - site_page: https://opendataloader.org/docs/rag-integration (fetched 2026-08-29T07:50:39.604473+00:00, sha 56ac896c5221)
- Data as of 2026-08-30T08:39:29.467469+00:00.