---
id: catboost
version: "1.2.10"
license: Apache License, Version 2.0
license_treatment: permissive
maintenance: active
---

# catboost — CatBoost Python Package

License: permissive · Maintenance: active · Downloads: 6.2M/mo

## What it is and what it does

CatBoost is a gradient boosting framework that builds ensembles of decision trees for supervised learning tasks. It is designed to handle both numerical and categorical features natively, reducing the need for manual feature engineering. The library includes GPU and multi-GPU training support, distributed training via Apache Spark, and built-in visualization tools for model interpretation.

The package depends on numpy, pandas, scipy, matplotlib, plotly, graphviz, and six. It ships as pre-compiled wheels for multiple Python versions and platforms (macOS, Linux, Windows). Installation is straightforward but medium-friction due to the compiled nature of the library. CatBoost is actively maintained with recent releases and is suitable for production use.

Use it for:

- Train gradient boosting models on tabular data with mixed numerical and categorical features without manual encoding
- Accelerate model training on GPU or multi-GPU systems for large datasets
- Build distributed gradient boosting pipelines using Apache Spark integration
- Generate feature importance rankings and visualizations for model interpretation
- Deploy pre-trained CatBoost models in production applications via the C++ API

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

CatBoost is a gradient boosting library that trains decision-tree ensembles for classification, regression, and ranking, with built-in support for categorical features and GPU acceleration.

Yes. CatBoost is production-stable, actively maintained, permissively licensed, and free of known vulnerabilities. Install it if you work with tabular data and want gradient boosting with native categorical feature support, GPU acceleration, or distributed training. The medium install friction is typical for compiled ML libraries and is not a barrier.

## Install

pip install catboost
uv add catboost
poetry add catboost

## Installing catboost

Before you install: Medium install friction due to compiled wheels for multiple Python versions and architectures. Active maintenance with recent releases; last commit 2026-08-14 and latest release 2026-02-18 indicate ongoing development. Depends on seven runtime packages including numpy, pandas, scipy, and visualization libraries.

License in practice: Licensed under Apache License, Version 2.0 (permissive). You can use, modify, and distribute CatBoost freely in commercial and private projects with minimal restrictions.

Quickstart:

pip install catboost

from catboost import CatBoostClassifier
import pandas as pd

X = pd.DataFrame({'feature': [1, 2]})
y = [0, 1]
model = CatBoostClassifier(verbose=0)
model.fit(X, y)

Verify before relying:

- Whether categorical feature auto-encoding works without preprocessing for all data types
- GPU support availability and setup requirements on different platforms
- Performance comparison claims against other GBDT libraries on your specific datasets
- Minimum Python version requirements and compatibility with Python versions beyond 3.14

## Package facts

- License: Apache License, Version 2.0 (permissive)
- Python support: unspecified
- Install friction: medium
- Maintenance: active
- Downloads: 6.2M/month (top 5,000 on PyPI)
- Known vulnerabilities: none known

## Tags

gradient boosting library, categorical feature handling, decision tree ensemble, GPU accelerated machine learning, catboost gradient boosting, fast gradient boosting, categorical data machine learning, gradient-boosting, categorical-features, gpu-accelerated

[View on SkillFed](https://skillfed.io/packages/catboost) · [View on PyPI](https://pypi.org/project/catboost/)