Author:	The Blosc development team
Contact:	[email protected]
Github:	https://github.com/Blosc/python-blosc2
Actions:
PyPi:
NumFOCUS:
Code of Conduct:

Python-Blosc2 is a high-performance compressor, compute engine, and format for binary data containers that are portable and open-source. It comes with a lazy expression engine allowing for complex calculations on compressed data, whether stored in memory, on disk, or over the network (e.g., via Caterva2). It is especially optimized for storing and retrieving data from N-dimensional arrays (NDArray) and columnar tables (CTable), complemented by a query/indexing layer. The main use case is fast, compressed, out-of-core numerical data — especially when data is too large to fit comfortably in RAM.

C-Blosc2 is used under the hood as its compression backend. Written in C, and building on its predecessor C-Blosc, C-Blosc2 aims to be an extremely fast meta-compressor for binary data, supporting a diverse set of strategies, and with an extensible plugin architecture for a wide range of codecs and filters.

miniexpr is embedded as the compute engine backend. It delivers fast multithreaded evaluation for lazy expressions directly over compressed chunks, and powers Menudet, a little language for portable, JIT-accelerated array (and table columns) kernels that can be serialized on a architecture-independent way.

More info: https://www.blosc.org/python-blosc2/getting_started/overview.html

Binary packages are available for major OSes (Win, Mac, Linux) and platforms. Install from PyPI using pip:

pip install blosc2 --upgrade

Conda users can install from conda-forge:

conda install -c conda-forge python-blosc2

Two CLI tools are installed along with the package:

• b2view: an interactive terminal browser (TUI) for TreeStore bundles (.b2d directories or .b2z files), with paged views of NDArray and CTable data of any size (walkthrough; requires pip install "blosc2[tui]").
• parquet-to-blosc2: converts Parquet files to Blosc2 columnar table stores, and back (walkthrough; requires pip install "blosc2[parquet]").

The documentation is available here:

https://blosc.org/python-blosc2/python-blosc2.html

You can find examples at:

https://github.com/Blosc/python-blosc2/tree/main/examples

A tutorial from PyData Global 2025 is available at:

https://github.com/Blosc/PyData-Global-2025-Tutorial

(Click here to watch the video recording of the tutorial)

It contains Jupyter notebooks explaining the main features of Python-Blosc2.

This software is licensed under a 3-Clause BSD license. A copy of the python-blosc2 license can be found in LICENSE.txt.

Discussion about this package is welcome at:

https://github.com/Blosc/python-blosc2/discussions

Stay informed about the latest developments by following us in Mastodon, Bluesky or LinkedIn.

Blosc2 is supported by the NumFOCUS foundation, the LEAPS-INNOV project and ironArray SLU, among many other donors. This allowed the following people to have contributed in an important way to the core development of the Blosc2 library:

• Francesc Alted
• Marta Iborra
• Luke Shaw
• Aleix Alcacer
• Oscar Guiñón
• Juan David Ibáñez
• Ivan Vilata i Balaguer
• Oumaima Ech.Chdig
• Ricardo Sales Piquer

In addition, other people have participated in the project in different aspects:

• Jan Sellner, contributed the mmap support for NDArray/SChunk objects.
• Dimitri Papadopoulos, contributed a large bunch of improvements to many aspects of the project. His attention to detail is remarkable.
• And many others that have contributed with bug reports, suggestions and improvements.

Developed using JetBrains IDEs.

You can cite our work on the various libraries under the Blosc umbrella as follows:

@ONLINE{blosc,
  author = {{Blosc Development Team}},
  title = "{A fast, compressed and persistent data store library}",
  year = {2009-2026},
  note = {https://blosc.org}
}

If you find Blosc useful and want to support its development, please consider making a donation or contract to the Blosc Development Team. Thank you!

Compress Better, Compute Bigger