chronos-bolt-tiny by Autogluon: Open-Weight Model

Runs On

What it takes to serve chronos-bolt-tiny (9M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

Precision	Cheapest setup	Per hour	Also fits
16-bit	1x MI300X (192 GB) Vultr	$1.85	1x H100 $1.99 · 1x MI325X $2.00
8-bit	1x MI300X (192 GB) Vultr	$1.85	1x H100 $1.99 · 1x MI325X $2.00
4-bit	1x MI300X (192 GB) Vultr	$1.85	1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

SAVRN's Notes on chronos-bolt-tiny

What do you forecast with 34.6 MB of weights? Power draw by the hour, utilization by rack, any series a facility keeps, and you do it without training anything, because Autogluon built the Chronos-Bolt family for zero-shot forecasting on a T5 encoder-decoder pretrained on nearly 100 billion observations. Its 9M parameters are stored as float32, so our memory table rounds every precision to 0.0 GB and names a 192 GB MI300X at $1.85 per hour on-demand because it has to name one.

Apache 2.0 permits commercial use, modification and redistribution if you keep the license, copyright and NOTICE files and state significant changes, plus an express patent grant. Our record has no context length for it, so the T5 and Chronos papers it cites are where you confirm how much history it accepts per forecast. Last updated October 30, 2025; no host on the SAVRN Index serves it.

SAVRN Research, 2026-09-18

Model Card

By Autogluon, published under apache-2.0, revision ea35bd742699.

Update Feb 14, 2025: Chronos-Bolt models are now available on Amazon SageMaker JumpStart! Check out the tutorial notebook to learn how to deploy Chronos endpoints for production use in a few lines of code.

Chronos-Bolt is a family of pretrained time series forecasting models which can be used for zero-shot forecasting. It is based on the T5 encoder-decoder architecture and has been trained on nearly 100 billion time series observations. It chunks the historical time series context into patches of multiple observations, which are then input into the encoder. The decoder then uses these representations to directly generate quantile forecasts across multiple future steps—a method known as direct multi-step forecasting. Chronos-Bolt models are more accurate, up to 250 times faster and 20 times more memory-efficient than the original Chronos models of the same size.

Performance

Read the full model card (636 words)

Configuration

Architecture: ChronosBoltModelForForecasting
Vocabulary size: 2
Stored precision: float32
Model type: t5

Identity and Version

Repository: autogluon/chronos-bolt-tiny
Publisher: Autogluon
Task: Time series forecasting
Modality: Time series
Library: Not stated by the source
Parameters: 9M parameters
Languages: Not stated by the source
Revision: ea35bd74269926879396f5c2476288aff9d76406
First published: 2024-11-13
Last updated: 2025-10-30

Files and Weights

4 files, 34.6 MB in total. The weights are 1 file totalling 34.6 MB in safetensors.

Weights1 file · 34.6 MB

Configuration1 file · 1.1 KB

Documentation1 file · 6.9 KB

Repository1 file · 1.5 KB

Every file

File	Type	Size	SHA-256
model.safetensors	Weights	34.6 MB	75068728d376
config.json	Configuration	1.1 KB	—
README.md	Documentation	6.9 KB	—
.gitattributes	Repository	1.5 KB	—

License and Download

License: apache-2.0
Access: Open weights, no gate
Download size: 34.6 MB

Download from Autogluon

Released by Autogluon through its official repository on Hugging Face. Read the license.

Built From

Described by arXiv:1910.10683
Described by arXiv:2403.07815

Memory Requirements

Precision	Weights in memory
As published	34.6 MB
16-bit	0.0 GB
8-bit	0.0 GB
4-bit	0.0 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About chronos-bolt-tiny

How much GPU memory does chronos-bolt-tiny need?

About 0 GB at 16-bit and 0 GB at 4-bit: the weights (9M parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run chronos-bolt-tiny on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use chronos-bolt-tiny commercially?

Yes. chronos-bolt-tiny is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.

Similar Models

Model · Time series forecasting

chronos-bolt-tiny

Amazon

Update Feb 14, 2025: Chronos-Bolt models are now available on Amazon SageMaker JumpStart! Check out the tutorial notebook to learn how to deploy Chronos endpoints for production use in a few lines of code. Chronos-Bolt is a family of pretrained time series forecasting models which can be used for zero-shot forecasting. It is based on the T5 encoder-decoder architecture and has been trained on nearly 100 billion time series observations. It chunks the historical time series context into patches of multiple observations, which are then input into the encoder. The decoder then uses these representations to directly generate quantile forecasts across multiple future steps—a method known as…

Open weights apache-2.0 9M parameters chronos-forecasting

View model

Model · Time series forecasting

chronos-t5-tiny

Amazon

Update Feb 14, 2025: Chronos-Bolt & original Chronos models are now available on Amazon SageMaker JumpStart! Check out the tutorial notebook to learn how to deploy Chronos endpoints for production use in a few lines of code. Update Nov 27, 2024: We have released Chronos-Bolt models that are more accurate (5% lower error), up to 250 times faster and 20 times more memory-efficient than the original Chronos models of the same size. Check out the new models here. Chronos is a family of pretrained time series forecasting models based on language model architectures. A time series is transformed into a sequence of tokens via scaling and quantization, and a language model is trained on these…

Open weights apache-2.0 8M parameters chronos-forecasting

View model

Model · Time series forecasting

chronos-t5-tiny

Autogluon

Update Feb 14, 2025: Chronos-Bolt & original Chronos models are now available on Amazon SageMaker JumpStart! Check out the tutorial notebook to learn how to deploy Chronos endpoints for production use in a few lines of code. Update Nov 27, 2024: We have released Chronos-Bolt models that are more accurate (5% lower error), up to 250 times faster and 20 times more memory-efficient than the original Chronos models of the same size. Check out the new models here. Chronos is a family of pretrained time series forecasting models based on language model architectures. A time series is transformed into a sequence of tokens via scaling and quantization, and a language model is trained on these…

Open weights apache-2.0 8M parameters transformers

View model

Model · Time series forecasting

flowstate

IBM Research

FlowState is the first time-scale adjustable Time Series Foundation Model (TSFM), open-sourced by IBM Research. Combining an State Space Model (SSM) Encoder with a Functional Basis Decoder allows FlowState to transition into a timescale invariant coefficient space and make a continuous forecast from this space. This allows FlowState to seamlessly adjust to all possible sampling rates. Therefore, training in one time-scale helps for inference at all scales, allowing for drastically improved utilization of training data across time-scales. This innovation leads to a significant improvement in performance, making FlowState the new state-of-the art in zero-shot time series forecasting.…

Open weights apache-2.0 9M parameters

View model

Model · Time series forecasting

granite-timeseries-flowstate-r1

IBM Granite

FlowState is the first time-scale adjustable Time Series Foundation Model (TSFM), open-sourced by IBM Research. Combining a State Space Model (SSM) Encoder with a Functional Basis Decoder allows FlowState to transition into a timescale invariant coefficient space and make a continuous forecast from this space. This allows FlowState to seamlessly adjust to all possible sampling rates. Therefore, training in one time-scale helps for inference at all scales, allowing for drastically improved utilization of training data across time-scales. This innovation leads to a significant improvement in performance, making FlowState the new state-of-the art in zero-shot time series forecasting.…

Open weights apache-2.0 9M parameters

View model

Model · Time series forecasting

moirai-2.0-R-small

Salesforce AI Research

Moirai 2.0 is a decoder-only universal time series forecasting transformer model pre-trained on: - Subset of GIFT-Eval Pretrain, and Train datasets (Non-leaking historical context). - Mixup data generated from non-leaking subsets of Chronos Dataset. - Synthetic time series produced via KernelSynth introduced in Chronos paper. - Internal Salesforce operational data. We make significant improvements over the first version of Moirai (please refer to the paper for previous version): - Switched from a distributional loss to a quantile loss formulation. - Moved from single-token to multi-token prediction, improving efficiency and stability. - Added a data filtering mechanism to filter out…

Open weights cc-by-nc-4.0 11M parameters

View model