StatsForecast
NixtlaNixtla's library of fast statistical forecasting models, compiled with numba and able to fit millions of series on Spark, Dask or Ray. It implements automatic ARIMA, ETS, CES and Theta selection, MSTL and TBATS for multiple seasonalities, GARCH, intermittent-demand models such as Croston and ADIDA, and baselines, about 35 models in all. Nixtla's closed TimeGPT forecasting service is a separate product.
Openness
5 high confidence- license
- Apache-2.0(OSI)
- source
- public(github.com/Nixtla/statsforecast)
- core features withheld
- no — Nixtla's paid offering is the separate TimeGPT service
Apache-2.0 and built from the public repository. Nixtla sells TimeGPT, a closed forecasting model reached through an API, and an enterprise plan built on it; the open libraries, StatsForecast included, are listed as free with nothing withheld.
- https://pypi.org/pypi/statsforecast/json recorded 2026-09-27
PyPI record for statsforecast 2.1.1, uploaded 2026-07-16, license "Apache Software License 2.0", Repository github.com/Nixtla/statsforecast.
- https://raw.githubusercontent.com/Nixtla/nixtla/HEAD/README.md recorded 2026-09-27
Nixtla SDK README: "TimeGPT itself is closed source. The SDK is a client".
- https://raw.githubusercontent.com/Nixtla/statsforecast/HEAD/LICENSE recorded 2026-09-27
LICENSE is the Apache License, Version 2.0, "Copyright 2022 Nixtla".
- https://ungh.cc/repos/Nixtla/statsforecast/files/main recorded 2026-09-27
Repository tree of 310 paths with no ee, enterprise, commercial or pro directory.
- https://www.nixtla.io/ recorded 2026-09-27
Nixtla pricing section: "Free Forever ... Open Source libraries: StatsForecast MLForecast NeuralForecast HierarchicalForecast UtilsForecast CoreForecast" and "Enterprise ... Everything in Open Source, plus: Premium Models and Updates Fine-tuned on your data".
Adoption
4 high confidenceMeasured on monthly PyPI downloads of the statsforecast package.
- https://pypistats.org/api/packages/statsforecast/recent recorded 2026-09-27
last_month 1,339,344 downloads of statsforecast
Capability
3 high confidenceStatsForecast is a catalog of statistical forecasters with the tools to fit and compare them across many series, and each Auto model selects its own settings. Choosing which model family to use is left to the user, as in sktime, whose toolkit is broader.
- https://raw.githubusercontent.com/Nixtla/statsforecast/HEAD/README.md recorded 2026-09-27
README: "Fastest and most accurate implementations of AutoARIMA, AutoETS, AutoCES, MSTL and Theta in Python"; models table covering Automatic Forecasting, ARIMA Family, Theta Family, Multiple Seasonalities, GARCH and ARCH, Baseline, Exponential Smoothing and Sparse or Intermittent models.
Verified 2026-09-27