What is Docling?

Docling is an open-source document processing tool that converts content from multiple formats into a unified structured representation. The solution supports PDF, DOCX, XLSX, PPTX, HTML, Markdown, images, audio files, and other formats, and performs advanced analysis of PDF documents, including reading order, tables, formulas, and layout elements.

Docling parses and exports documents into Markdown, HTML, WebVTT, DocTags, JSON, and other formats without losing structural semantics. This makes it suitable for building retrieval-augmented generation (RAG) pipelines, semantic search, document indexing, text extraction, and preparing datasets for further machine analysis.

Additionally, Docling supports local execution mode for sensitive data and isolated environments, as well as optical character recognition (OCR) for scanned PDFs and images. This enables its use not only in open integration scenarios but also in environments where document processing must remain within controlled infrastructure boundaries.