Chronos-2 and Chronos-Bolt: Pretrained Time Series Forecasters from Amazon Science
Chronos: Pretrained Models for Time Series Forecasting
At a glance
- What is it?
- Amazon's chronos-forecasting package wraps pretrained models that forecast time series without task-specific training. This review covers the architecture, the zero-shot claims, the deployment paths, and the limits of trusting a foundation model with your data.
- Who is it for?
- Adopt chronos-forecasting if you need zero-shot forecasts on univariate or multivariate series, especially with covariates, and you can accept the Apache-2.0 license and the Hugging Face download requirement. Skip it if your series have strong non-stationary regimes or if you need on-premise, offline inference without external model pulls.
- Can I use it commercially?
- Yes. Apache-2.0 is a permissive licence: you can use, modify and sell software built on it, as long as you keep its copyright and licence notices.
- Is it still maintained?
- Yes. The repository last received commits 41 days ago.
- What is it written in?
- Mainly Python, according to GitHub's language statistics.
Answers come from the project's GitHub data, last synced on November 15, 2026, and from our analysis. They are not legal advice.
Star history
- amazon-science/chronos-forecasting6,310 ★ · +470
| Repository | 2026-09-10 | 2026-11-15 |
|---|---|---|
| amazon-science/chronos-forecasting | 5.8k | 6.3k |
470 Stars in the last 66 days, measured from the daily GitHub snapshot. Compare with other projects
Editorial analysis
What Chronos Actually Solves
Most time series forecasting tools require you to train a model per dataset. You collect history, engineer features, tune hyperparameters, and retrain when the pattern drifts. Chronos takes a different route: it ships pretrained models that you point at a series and ask for quantiles. The package targets engineers who want a forecast without building a bespoke model, or who need a strong baseline before investing in custom work. The README positions Chronos-2 as the latest model with zero-shot support for univariate, multivariate, and covariate-informed tasks. That means it accepts external regressors. The intended user is someone comfortable with Python and Hugging Face model hubs, not a person looking for a point-and-click tool.
Three Model Families, Three Different Mechanisms
The repository exposes three generations of models, each with a distinct inference strategy. The original Chronos treats a time series like text: it scales and quantizes observations into tokens, then feeds them to a language model trained with cross-entropy loss. Forecasts come from sampling multiple future trajectories. Chronos-Bolt changes the input representation. It patches the historical context into chunks of several observations, passes those patches through an encoder, and has the decoder directly generate quantile forecasts for multiple future steps. This direct multi-step approach is why the README claims Bolt models are up to 250 times faster and 20 times more memory efficient than the original Chronos models of the same size. Chronos-2, the newest family, adds zero-shot support for univariate, multivariate and covariate-informed forecasting. The README reports a win rate over 90% against Chronos-Bolt in head-to-head comparisons, with the largest gains on tasks with exogenous features. These are architectural differences, not just version bumps.
Getting It Running: Commands and a Minimal Example
Installation is a single pip command: pip install chronos-forecasting. The README's minimal example imports Chronos2Pipeline from chronos, loads the model with Chronos2Pipeline.from_pretrained("amazon/chronos-2", device_map="cuda"), and reads context and future covariate data from Parquet files with pandas. What is visible is that you need pandas with the pyarrow extra, installed via pip install 'pandas[pyarrow]'. For a fuller walkthrough, the README points to the notebooks/chronos-2-quickstart.ipynb notebook. Model weights are not bundled in the pip package. They are downloaded from Hugging Face by model ID, like amazon/chronos-2 or autogluon/chronos-2-small. That means your inference environment needs network access to huggingface.co on first use, and you must pin a model ID to get reproducible results. The package also lists a SageMaker deployment notebook, which suggests that for production you are expected to move off a local Python script and onto an AWS endpoint.
The Zero-Shot Promise and Its Boundaries
Zero-shot forecasting sounds magical, but the README's own claims reveal the boundaries. Chronos-2's benchmark wins are on fev-bench, GIFT-Eval, and Chronos Benchmark II, which are public evaluation suites. Those benchmarks do not cover every domain. If your series has a strong weekly seasonality and a holiday effect that the pretrained model never saw, the forecast may be off. Check the context length and minimum series length of the model you pick before relying on it for very long or very short series.
Where Chronos Is the Wrong Tool
Chronos is built around zero-shot forecasting from pretrained weights. If you want a model fitted only to your own historical records, a statistical model like ARIMA or a gradient boosting library trains on your series directly. Chronos also assumes you can tolerate the latency of a transformer model. Even the tiny Bolt model has 9 million parameters, and the base model has 205 million. On a CPU-only environment, inference for long horizons could be slow. The README suggests SageMaker for production, which implies that running these models at scale on your own hardware may require GPU instances. If your deployment is an embedded device or a low-power edge node, a 200-million-parameter model is impractical.
The Real Alternative: Train Your Own or Use a Statistical Baseline
The honest alternative to Chronos is not another pretrained model. It is a classical method like seasonal naive or exponential smoothing, which you can implement with statsmodels or a few lines of pandas. Those methods have no pretrained weights, no Hugging Face download, and no GPU requirement. They also give you interpretable parameters: you can see the level, trend, and seasonality components. The trade-off is accuracy on complex series. Chronos-2's benchmark results suggest it beats those baselines on public datasets, but a baseline is still the right first check. If a seasonal naive forecast is within 5% of Chronos's error on your data, you do not need a 120-million-parameter model. Another alternative is AutoGluon, which the README mentions for deployment. AutoGluon is a full AutoML framework that can train multiple model types and stack them. It learns from your data rather than relying only on pretraining. The difference in approach is fundamental: Chronos generalizes from pretraining, AutoGluon fits to your data.
Maintenance, Licensing, and Upgrade Path
The repository is under the Apache-2.0 license, which permits commercial use, modification, and redistribution, with the condition that you preserve the license notice. That is permissive; the model weights are downloaded from Hugging Face, so check the terms on each model page as well. The project is actively maintained: the last push was on 2026-10-05, and version 2.3.2 was released on 2026-09-08. The release history shows a steady cadence, with 2.3.0 in June and 2.3.1 in July 2026. That means you can expect updates, but also that the API might shift between minor versions. If you pin to a specific version, you are safe, but if you follow the latest release, you should read the changelog before upgrading. The package also depends on the Hugging Face transformers library, which has its own release cycle. Upgrading transformers to get a new feature could break Chronos, so you should test after any dependency bump.
Editorial conclusion
Adopt chronos-forecasting if you need zero-shot forecasts on univariate or multivariate series, especially with covariates, and you can accept the Apache-2.0 license and the Hugging Face download requirement. Skip it if your series have strong non-stationary regimes or if you need on-premise, offline inference without external model pulls. Before committing, verify the model's context length matches your series length, test on your own metrics, and compare against a simple baseline like seasonal naive. For production, review the SageMaker deployment notebooks and confirm the CPU or GPU instance type you plan to use meets the memory footprint of the selected model size.
Frequently asked questions
What is chronos forecasting?
It is a Python package providing an interface to the Chronos family of pretrained time series forecasting models, licensed under Apache-2.0 and installable with pip install chronos-forecasting. The models are downloaded from a model hub rather than shipped inside the package, and inference runs locally on torch.
How does Chronos-2 perform time series forecasting?
It offers zero-shot support for univariate, multivariate and covariate-informed tasks, and the project reports state-of-the-art zero-shot performance across benchmarks including fev-bench and GIFT-Eval, with the largest improvements on tasks that include exogenous features. It also reports a win rate of over 90 percent against Chronos-Bolt in head-to-head comparisons.
Which models are available through chronos-forecasting?
Twelve model IDs across three families. Chronos-2 has amazon/chronos-2 at 120M plus autogluon/chronos-2-synth at 120M and autogluon/chronos-2-small at 28M; Chronos-Bolt has tiny 9M, mini 21M, small 48M and base 205M; the original t5 line has tiny 8M, mini 20M, small 46M, base 200M and large 710M.
What are the Python and dependency requirements for chronos-forecasting?
Python 3.10 or later, with six runtime dependencies that all carry upper bounds: torch>=2.2,<3, transformers>=4.41,<6, accelerate>=1.1.0,<2, numpy>=1.21,<3, einops>=0.7.0,<1 and pandas>=2.0,<4. The build backend is hatchling and the wheel packages src/chronos.
Does Amazon recommend running chronos-forecasting in production as a library?
The README recommends deploying Chronos-2 to Amazon SageMaker for production use rather than depending on the pip package. Two options are named: AutoGluon-Cloud as a high-level Python API with real-time, serverless and batch inference, and SageMaker JumpStart for production-ready real-time endpoints on CPU or GPU.
Official sources
Add this badge to your README
If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.
[](https://hysenlabs.com/projects/amazon-science-chronos-forecasting)Star history chart for your README
[](https://hysenlabs.com/projects/amazon-science-chronos-forecasting)Was this useful?