---
id: autogluon-timeseries
version: "1.6.1"
license: Apache-2.0
license_treatment: permissive
maintenance: active
---

# autogluon.timeseries — Fast and Accurate ML in 3 Lines of Code

License: permissive · Maintenance: active · Downloads: 347.3K/mo

## What it is and what it does

AutoGluon TimeSeries is an automated machine learning library that simplifies time series forecasting by automatically selecting, training, and ensembling models from deep learning, statistical, and foundation model approaches. It wraps complex model selection and hyperparameter tuning into a few lines of code, targeting developers who want strong forecasting performance without manual model engineering.

The package depends on a large ecosystem: PyTorch and Lightning for neural networks, Hugging Face transformers and PEFT for foundation models, GluonTS and specialized forecasting libraries (statsforecast, mlforecast, chronos-forecasting) for statistical and hybrid approaches, and AutoGluon's own tabular and feature modules for preprocessing and ensemble logic. It is actively maintained, supports Python 3.10–3.13, and runs on Linux, macOS, and Windows.

Use it for:

- Build a demand forecasting model for supply chain without manually tuning ARIMA, Prophet, or neural network hyperparameters.
- Forecast financial time series (stock prices, trading volumes) by combining statistical and deep learning models automatically.
- Generate probabilistic forecasts with uncertainty intervals for risk assessment in healthcare or operations planning.
- Rapidly prototype multiple forecasting approaches on new datasets to identify which model class performs best.
- Deploy production forecasting pipelines that adapt to new data by retraining with minimal code changes.

## Worth the install?

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

AutoGluon TimeSeries automates machine learning for time series forecasting, training and deploying high-accuracy models with minimal code using deep learning and statistical approaches.

Yes, with conditions. Install if you need time series forecasting with minimal manual model selection and have the disk space and compute for PyTorch and 27 dependencies. The active maintenance, permissive license, low install friction, and zero known vulnerabilities make it a solid choice. Skip if you need lightweight, dependency-minimal forecasting or are constrained to Python <3.10.

## Install

pip install autogluon-timeseries
uv add autogluon-timeseries
poetry add autogluon-timeseries

## Installing autogluon.timeseries

Before you install: Low install friction with a pure-Python wheel. Active maintenance with a recent release (8 days old) and strong repository signals (10596 stars). Requires 27 runtime dependencies including PyTorch, Lightning, Hugging Face transformers, and specialized forecasting libraries, which may take time to resolve but are well-established packages.

License in practice: Apache-2.0 permissive license allows use in commercial and private projects with minimal restrictions, requiring only license and copyright notice retention.

Quickstart:

pip install autogluon.timeseries

from autogluon.timeseries import TimeSeriesPredictor
predictor = TimeSeriesPredictor(label="target").fit(train_data, presets="best")
predictions = predictor.predict(test_data)

Requires Python 3.10–3.13; PyTorch and other ML dependencies will download on first install, which may require several gigabytes of disk space.

Verify before relying:

- Whether the package includes GPU acceleration support or if that requires separate installation steps.
- Specific performance benchmarks or accuracy claims for different forecasting scenarios.
- Memory and compute requirements for typical time series datasets.

## Package facts

- License: Apache-2.0 (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 347.3K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

time series forecasting automation, automl for forecasting, deep learning time series, probabilistic forecasting library, automated model selection forecasting, time series prediction, ensemble forecasting models, automl, time-series-forecasting, ensemble-learning

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