dmlc XGBoost
XGBoost is an open-source gradient boosting library maintained by the Distributed Machine Learning Community (dmlc).
Profile
XGBoost is an open-source software library that provides a fast and scalable implementation of gradient boosting for machine learning, primarily used for classification and regression on structured data.
XGBoost is an open-source gradient boosting library maintained by the Distributed Machine Learning Community (dmlc). It was created in 2014 by Tianqi Chen, then a PhD student at the University of Washington, and has since become one of the most widely used machine learning algorithms in both industry and competition settings. As of June 2026, the project has over 28,400 stars and 8,900 forks on GitHub, with 7,834 commits and 76 tags.
The library is written in C++ but provides interfaces for Python, R, Java, Scala, Julia, and other languages, making it accessible across the data science ecosystem. XGBoost is known for its speed and performance, particularly on structured/tabular data, and has been a dominant algorithm on platforms like Kaggle. The project is funded through Open Collective and GitHub Sponsors, with no disclosed venture capital funding or corporate ownership.
The most recent commit as of June 5, 2026, was by contributor trivialfis, who added faster GPU sampling using the Philox algorithm. XGBoost does not have a formal company structure, headquarters, or paid employees; it is maintained by volunteers and a small group of core contributors. The library's financial model relies on donations and sponsorships, with no public revenue figures.
Despite its lack of corporate backing, XGBoost is embedded in production systems at major technology firms, including NVIDIA, which uses it in its RAPIDS suite for GPU-accelerated machine learning. The project's governance is informal, with decisions made by a small group of maintainers. There is no known customer churn, leadership turnover, or controversy, but the project faces risks from competing gradient boosting libraries like LightGBM and CatBoost, as well as the rise of deep learning for tabular data. XGBoost's long-term sustainability depends on continued community contributions and sponsorship.
Who buys this
- Data scientists and machine learning engineers building predictive models for structured data
- Enterprises using machine learning for fraud detection, credit scoring, and customer churn prediction
- Academic researchers and students in machine learning and data science courses
- Competition participants on platforms like Kaggle and DrivenData
- Cloud and hardware vendors integrating XGBoost into their AI platforms (e.g., NVIDIA RAPIDS, AWS SageMaker, Google Cloud AI)
Strengths and what to watch
Strengths
- Proven performance: XGBoost has been the winning algorithm in numerous Kaggle competitions and is widely cited in academic papers for its accuracy on tabular data.
- Scalable and efficient: The library supports distributed training on CPUs and GPUs, with recent optimizations like Philox-based sampling for faster GPU execution.
- Broad ecosystem support: XGBoost integrates with major programming languages (Python, R, Java, Julia) and cloud platforms, making it a default choice for many production pipelines.
Watch for
- Competition from other gradient boosting libraries: LightGBM and CatBoost have gained significant adoption, and XGBoost's relative performance advantage is narrowing.
- Dependence on volunteer maintainers: The project has no paid staff or formal governance, which could lead to slower bug fixes or feature development compared to commercially backed alternatives.
- Shift to deep learning for tabular data: New architectures like TabNet and FT-Transformer are challenging the dominance of gradient boosting on structured data, potentially reducing XGBoost's relevance.
Recent moves
Key Information
- Industry
- AI Frameworks, Tools & Libraries
- Founded
- 2014
Frequently Asked Questions
What is XGBoost and what is it used for?
XGBoost is an open-source gradient boosting library for machine learning. It provides a fast and scalable implementation of gradient boosting, primarily used for classification and regression on structured data. It was created in 2014 by Tianqi Chen.
Is XGBoost still relevant in 2026?
Yes, XGBoost remains widely used in industry and competitions. As of June 2026, it has over 28,400 GitHub stars and is embedded in production systems at major firms like NVIDIA. However, it faces competition from LightGBM, CatBoost, and deep learning for tabular data.
How does XGBoost compare to LightGBM and CatBoost?
XGBoost is known for its proven performance and scalability, but its relative advantage over LightGBM and CatBoost is narrowing. All three are gradient boosting libraries, but XGBoost has broader ecosystem support and a longer track record in competitions like Kaggle.
What programming languages does XGBoost support?
XGBoost is written in C++ but provides interfaces for Python, R, Java, Scala, Julia, and other languages. This makes it accessible across the data science ecosystem and easy to integrate into various production pipelines.
Who maintains XGBoost and how is it funded?
XGBoost is maintained by volunteers and a small group of core contributors from the Distributed Machine Learning Community. It has no paid employees or formal company structure. Funding comes from donations via Open Collective and GitHub Sponsors.
What are the main risks facing XGBoost?
Key risks include competition from LightGBM and CatBoost, dependence on volunteer maintainers without paid staff, and the rise of deep learning architectures for tabular data like TabNet. These could slow development or reduce XGBoost's relevance over time.
Sources
- github.com — GitHub repository metadata: stars, forks, commits, tags, and recent commit details.
- nvidianews.nvidia.com — NVIDIA's fiscal 2026 revenue and data center revenue, indicating scale of AI infrastructure that may use XGBoost.
- futurumgroup.com — NVIDIA Q3 FY 2026 results and Blackwell GPU adoption, relevant to XGBoost's GPU acceleration context.
- techcrunch.com — TechCrunch article on 2025 unicorn creation, providing context for the startup funding environment around XGBoost.