Skip to content

Features

Kreuzberg is a document intelligence library built on a Rust core with bindings for 12 languages. It extracts text, tables, and metadata from 91+ file formats, runs OCR when needed, and feeds the results through a configurable post-processing pipeline. Each capability below links to the guide or reference page with configuration details and code examples.

Capabilities

Section titled “Capabilities”
  • Extraction — Pull text, tables, metadata, and images from 91+ formats through native Rust extractors. No external tools such as LibreOffice are required. See Extraction Basics.
  • OCR — Recognize text in scanned documents and images with Tesseract, PaddleOCR, or EasyOCR, optionally chained into a quality-driven fallback pipeline. See the OCR Guide.
  • Chunking and embeddings — Split extracted text into sized chunks and generate local vector embeddings with FastEmbed for RAG pipelines. See Chunking and Embeddings.
  • Output formats — Render results as plain text, Markdown, Djot, HTML, or structured JSON. The HTML renderer ships styled kb-* classes and built-in themes. See Output Formats and HTML Output.
  • API and MCP server — Serve extraction over HTTP (kreuzberg serve) or expose it to AI agents over MCP (kreuzberg mcp). See the API Server Guide.
  • Plugins — Extend the pipeline with custom document extractors, OCR backends, validators, and post-processors. See the Plugin System concept and the Creating Plugins guide.

Format Support

Section titled “Format Support”

Kreuzberg handles 91+ file formats across these categories:

  • Documents — PDF, Word, PowerPoint, OpenDocument, RTF, plain text, Markdown, and more
  • Spreadsheets — Excel, Numbers, OpenDocument, CSV, TSV, dBASE
  • Images — JPEG, PNG, GIF, BMP, TIFF, WebP, JPEG 2000, and more (routed to OCR)
  • Email — EML, MSG
  • Web and markup — HTML, XHTML, XML, SVG
  • Structured data — JSON, YAML, TOML
  • Archives — ZIP, TAR, GZIP, 7-Zip
  • Academic — EPUB, BibTeX, RIS, LaTeX, Typst, JATS, DocBook, and more

For the full format matrix with MIME types and extraction methods, see the Format Support Reference.

Pipeline

Section titled “Pipeline”

Every file flows through the same multi-stage pipeline: MIME detection, format extraction, optional OCR, post-processing (validators, quality processing, chunking, embeddings), and caching. For a stage-by-stage walkthrough, see Extraction Pipeline.

Deployment Modes

Section titled “Deployment Modes”
Mode When to Use Details
Library Embedding extraction into your application Import the package in Python, TypeScript, Rust, Go, Ruby, C#, Java, PHP, Elixir, R, or C
CLI One-off extractions, scripting, CI pipelines kreuzberg extract document.pdf --format json — see CLI Usage
REST API Multi-service architectures, language-agnostic access kreuzberg serve --port 8000 — see API Server Guide
MCP Server AI agent integration kreuzberg mcp — stdio transport with JSON-RPC 2.0
Docker Reproducible deployments with all dependencies bundled ghcr.io/kreuzberg-dev/kreuzberg-full:4 — see Docker Guide

Configuration

Section titled “Configuration”

Kreuzberg reads configuration programmatically or from kreuzberg.toml, kreuzberg.yaml, or kreuzberg.json, auto-discovered from the current directory, ~/.config/kreuzberg/, and /etc/kreuzberg/. Environment variables override file-based settings. See the Configuration Guide.

Next Steps

Section titled “Next Steps”