AutoGluon

AutoGluon is AWS’s open-source AutoML library that delivers fast, accurate machine learning in as little as three lines of code. Point it at your data and it automatically trains, tunes, and ensembles strong models for tabular prediction, image classification, object detection, text tasks, and time-series forecasting — routinely reaching leaderboard-level results without manual feature engineering.

Overview

Rather than betting on a single algorithm, AutoGluon trains a diverse portfolio of models — from gradient-boosted trees to deep neural networks — and stacks them into ensembles that are typically far more accurate than any individual model. Its tabular module is a favorite of Kaggle competitors and data teams alike, handling messy real-world columns, missing values, and mixed data types automatically. Dedicated modules cover vision, text, multimodal data, and forecasting, all sharing the same compact API. The project has more than 10,700 GitHub stars, is under active development with updates in October 2026, and is maintained by the Amazon Science and AWS AI communities under the Apache-2.0 license.

Key Features

• State-of-the-art results on tabular data with a three-line API
• Automatic ensembling and stacking of many model types
• Modules for images, text, multimodal data, and time-series forecasting
• Robust handling of messy data, missing values, and mixed column types
• Configurable time budgets — quick prototypes or long, thorough searches
• Battle-tested in Kaggle competitions and production data teams

💳 Pricing

AutoGluon is completely free and open source under the Apache-2.0 license. You only need your own compute for training.

Who Is It For

Data scientists and developers who want strong predictive models quickly — for tabular business data, forecasting, or as a powerful baseline for any ML task.