Skip to content

TiRex-2

Generalizing TiRex to Multivariate Data and Streaming

Paper Hugging Face GitHub PyPI Docker Open in Colab TiRex-2 Demo

🎉 News (Sep 2026)

TiRex-2 has been accepted as an Oral presentation (0.36%) at NeurIPS 2026!

TiRex-2 is a pretrained time series foundation model that forecasts one or many target variates directly from their history, optionally conditioned on past and future-known covariates. A single checkpoint serves both univariate and multivariate forecasting, built on a recurrent architecture designed for efficient streaming settings — all zero-shot, with no task-specific training or fine-tuning.

TiRex-2 generalizes the original univariate model, TiRex, to multivariate forecasting with past and future covariates. See the Introduction for background and the paper for details.

Looking for TiRex-1?

This site documents TiRex-2. Documentation for the original univariate TiRex model lives at nx-ai.github.io/tirex — the two are separate projects and this site does not attempt to unify them.

Key facts

  • Zero-shot multivariate forecasting — TiRex-2 forecasts multiple target variates out of the box, without training or fine-tuning on your data.
  • Past and future-known covariates — TiRex-2 natively conditions on past covariates and future-known covariates, such as calendar features, holidays, promotions, or scheduled interventions.
  • Small active footprint — TiRex-2 activates 38.4M parameters in univariate mode and an additional 44.1M parameters for multivariate forecasting.

Where to go next

  • Installation — pip install tirex-2, Pixi setup, and gated Hugging Face weight access.
  • Quickstart — minimal sine-wave and covariate examples.
  • How-to guides — univariate/multivariate forecasting, covariates in depth, and what streaming does (and doesn't) mean in this open-source release.
  • Deployment — the Docker-based HTTP/MQTT/MCP inference server.
  • API reference — generated reference for the public tirex2 API.

TiRex-2 PRO

This repository contains NXAI's open-source release of TiRex-2. TiRex-2 PRO extends it with:

  • Support: dedicated support and engagement from NXAI.
  • Streaming: incremental forecast updates as new observations arrive, without recomputing forecasts over the full history.
  • Speed: optimized inference, including hardware-specific optimizations for edge, embedded, and industrial PC deployments.
  • Fine-tuning: models adapted to your data or pretrained on different datasets for improved performance.
  • Classification & Regression: TiRex-2 adapted for classification and regression tasks.

See TiRex-2 PRO for the full feature comparison, or contact contact@nx-ai.com.

Cite Our Work

If you use TiRex-2 in your research, please cite our work:

@misc{podest2026tirex2generalizingtirexmultivariate,
      title={TiRex-2: Generalizing TiRex to Multivariate Data and Streaming},
      author={Patrick Podest and Marco Pichler and Elias Bürger and Levente Zólyomi and Bernhard Voggenberger and Wilhelm Berghammer and Daniel Klotz and Sebastian Böck and Günter Klambauer and Sepp Hochreiter},
      year={2026},
      eprint={2607.01204},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2607.01204},
}

Intended use and limitations

This release supports time-series forecasting as documented on this site. High-risk applications under EU AI Act Article 6 are outside NXAI's intended purpose. NXAI has not validated this release for high-risk compliance. Users and integrators must assess suitability and meet their legal obligations. Unlawful uses, including those prohibited by EU AI Act Article 5, are not permitted. See the Intended Use Notice for details. It does not change the licence or statutory obligations.

License

TiRex-2 is licensed under the Apache License 2.0.