chonkie Releases

A major release introducing HTML table support via TableChef and TableChunker, a self-hostable FastAPI chunking server, native async capabilities, and a migration to the `ty` type-checking library.

Mar 11, 2026

This release extends the Visualizer to support lists of strings and includes various dependency updates, performance improvements, and bug fixes.

Feb 16, 2026

Focuses on advanced recognition capabilities with the new Language Models integration (LangExtract) and GPU optimizations. Strengthens Docker security with non-root users and healthchecks, adding recognizers for Thai IDs and Indian GST.

Feb 12, 2026

This foundational release establishes core infrastructure including Docker and MPS support, adds Apache licensing, introduces LM selection and MLX Gemma support, and adds support for various STT/TTS models like Paraformer and ChatTTS.

Feb 6, 2026

A patch release focused on stability, fixing OpenAI lazy imports, multiprocessing safety, and edge cases in CodeChunker and SlumberChunker.

Feb 4, 2026

This release introduces the Chonkie CLI and completes the migration from Cython to Rust for improved performance and maintainability.

Jan 28, 2026

Enhances CSV parsing error diagnostics with verbose modes and fixes date format detection issues in the Tally project.

Jan 17, 2026

Adds CSV export functionality, dynamic transaction description transformation, and improves supplemental data source handling.

Jan 14, 2026

Introduces directory and glob pattern support for CSV sources, improves European CSV handling, and refines tag collection logic.

Jan 10, 2026

Replaces the primary command with `tally up`, adds a report diff feature for tracking changes, and introduces experimental supplemental data sources.

Jan 6, 2026

Adds a new `FastChunker` implementation to significantly improve chunking speed, along with minor type safety improvements.

Jan 5, 2026

A maintenance release focused on compatibility and code quality, adding Python 3.14 wheel support, enforcing Ruff formatting, and fixing dependency constraints for Weaviate tests.

Dec 26, 2025

Introduces new tokenization capabilities and Python version support, dropping Python 3.9 support while adding Python 3.13 support.

Dec 16, 2025

A bug-fixing release that resolves issues with Azure OpenAI embeddings and file fetching, while replacing the loguru dependency with standard logging.

Nov 14, 2025

Adds support for `MilvusHandshake` for vector store integration and fixes various documentation and import issues.

Nov 3, 2025

A major feature release introducing `chonkie.Pipeline` for composable workflows, `TableChef` support, and new visualizer themes.

Oct 11, 2025

Adds Markdown support and introduces table chunking capabilities with a refactor of the Document class.

Sep 27, 2025

This release introduces search methods to handshakes and file input for cloud chunkers, alongside a major breaking change that standardizes return types to the generic `Chunk` class.

Sep 19, 2025

Introduces a new Korean Resident Registration Number recognizer and integrates Azure Health Data Services (AHDS) for medical de-identification. Includes the Slim NLP Engine, comprehensive YAML validation, and security patches.

Sep 9, 2025

A maintenance release focused on fixing compilation flags to ensure compatibility across different Linux distributions.

Aug 28, 2025