Mlpack

mlpack is a free, open-source, header-only C++ machine learning library with bindings for Python, Julia, R, Go, and the command line, fiscally sponsored by NumFOCUS and backed by Georgia Tech, Freie Universität Berlin, and Google Summer of Code. It's built on established libraries (Armadillo, ensmallen, cereal) for fast, extensible algorithm implementations, and its design has been validated through peer review (a 2023 Journal of Open Source Software paper) rather than only marketing claims. Being header-only C++ means it's aimed squarely at developers comfortable with systems-level programming who need ML performance in a compiled-language context, not a general audience. It's completely free under the permissive 3-clause BSD license, with the project accepting donations through NumFOCUS to fund ongoing development. The current release (4.8.0, June 2026) shows active, continued maintenance. Good fit for a developer who needs fast, embeddable machine learning in a C++ application, or wants ML bindings across multiple languages from one shared codebase.