kreuzberg-dev/kreuzberg v4.1.0

Release v4.1.0

on GitHub

[4.1.0] - 2026-01-21

Added

API

• POST /chunk endpoint: New text chunking endpoint for breaking text into smaller pieces
  • Accepts JSON body with text, chunker_type (text/markdown), and optional config
  • Returns chunks with byte offsets, indices, and metadata
  • Configuration options: max_characters (default: 2000), overlap (default: 100), trim (default: true)
  • Supports both text and markdown chunking strategies

Core

• Djot markup format support: New .djot file extraction with comprehensive Djot syntax support

  • Full parser implementation with structured representation via DjotContent type
  • Feature-gated behind djot feature flag (enabled by default)
  • 39 comprehensive tests covering Unicode, tables, roundtrip conversion, and edge cases

• Content output format configuration: New ContentFormat enum for configurable text output formatting

  • Converts extracted content from ANY file format to Plain, Markdown, Djot, or HTML
  • CLI support with --content-format flag and KREUZBERG_CONTENT_FORMAT environment variable

• Element-based output format: New ResultFormat::ElementBased option provides Unstructured.io-compatible semantic element extraction

  • Extracts structured elements: titles, paragraphs, lists, tables, images, page breaks, headings, code blocks, block quotes, headers, footers
  • Each element includes rich metadata: bounding boxes, page numbers, confidence scores, hierarchy information

Language Bindings

• All language bindings (Python, TypeScript, Ruby, PHP, Go, Java, C#, Elixir, WASM) updated with:
  • Content format configuration support
  • Result format configuration for element-based output
  • DjotContent and element types

Documentation

• Djot format documentation: New format reference and usage examples
• Migration guides: New documentation for Unstructured.io users

Changed

Codebase

• Major refactoring for maintainability: Split 22 large monolithic files into 110+ focused modules

Fixed

CI/CD

• Ruby macOS builds: Fixed unused imports causing compilation failures
• TypeScript tests on ARM64: Fixed module resolution error
• Go Windows builds: Disabled incompatible verbose linker flags
• PHP Windows builds: Added documentation for cargo fingerprint cache corruption issues

Documentation

• MkDocs build: Fixed broken benchmark documentation links

Python

• Type exports: Fixed missing type exports in kreuzberg.types.__all__

Elixir

• DOCX keyword extraction: Fixed FunctionClauseError when extracting DOCX files with keywords metadata

See CHANGELOG.md for full details.