The Best 52 Climate Model Tools in 2025

Chronos is a family of pre-trained time series forecasting models based on language model architectures. It converts time series into token sequences through quantization and scaling for training, suitable for probabilistic forecasting tasks.

Chronos-Bolt is a series of pretrained time series foundation models based on the T5 architecture, achieving efficient time series forecasting through innovative chunk encoding and direct multi-step prediction

Chronos-Bolt is a series of pretrained time series forecasting models that support zero-shot prediction with high accuracy and fast inference speed.

Chronos-Bolt is a series of pre-trained time series forecasting models that support zero-shot prediction, based on the T5 encoder-decoder architecture, trained on nearly 100 billion time series observations.

Chronos is a family of pre-trained time series forecasting models based on language model architecture, which transforms time series into token sequences for training to achieve probabilistic forecasting.

TinyTimeMixers (TTMs) are compact pre-trained models for multivariate time series forecasting open-sourced by IBM Research, with fewer than 1 million parameters, excelling in zero-shot and few-shot forecasting tasks.

Chronos is a family of pretrained time series forecasting models based on language model architectures, trained by quantizing and scaling time series into token sequences.

Chronos-Bolt is a series of pretrained time series forecasting models supporting zero-shot prediction, based on T5 encoder-decoder architecture, trained on nearly 100 billion time series observations.

This is a pre-trained time series forecasting model, focusing on time series data analysis and prediction tasks.

Chronos-Bolt is a series of pretrained time series forecasting models capable of zero-shot prediction, based on the T5 encoder-decoder architecture, trained on nearly 100 billion time series observations.

Chronos-Bolt is a series of pretrained time series forecasting models supporting zero-shot prediction, based on T5 encoder-decoder architecture, trained on nearly 100 billion time series observations.

Chronos-Bolt is a series of pretrained time series forecasting models supporting zero-shot forecasting, based on T5 encoder-decoder architecture, trained on nearly 100 billion time series data points.

Chronos is a family of pretrained time series forecasting models based on language model architectures, trained by quantizing and scaling time series into token sequences.

TinyTimeMixers (TTMs) are compact pretrained models for multivariate time-series forecasting open-sourced by IBM Research, starting from 1 million parameters, introducing the concept of 'miniature' pretrained models in time-series forecasting for the first time.

Chronos is a family of pretrained time series forecasting models based on language model architecture, which converts time series into token sequences through quantization and scaling for training, supporting probabilistic forecasting.

A PyTorch-based pretrained model for time series forecasting, suitable for various time series data analysis tasks

Moirai is a general-purpose time series forecasting Transformer based on a masked encoder, pre-trained as a large time series model on the LOTSA dataset.

Chronos is a family of pre-trained time series forecasting models based on language model architectures, which transform time series into token sequences for training through quantization and scaling.

Chronos-Bolt is a series of pretrained time series forecasting models supporting zero-shot prediction, based on the T5 encoder-decoder architecture, trained on nearly 100 billion time series observations.

Chronos-T5 is a pretrained time series forecasting model based on a language model architecture, achieving probabilistic forecasting by converting time series into token sequences and training the language model using cross-entropy loss.

Chronos is a family of pretrained time series forecasting models based on language model architecture, which supports probabilistic forecasting by converting time series into token sequences for training.

Chronos-T5 is a pre-trained time series forecasting model based on a language model architecture. It converts time series into token sequences through quantization and scaling for training, making it suitable for various time series forecasting tasks.

Moirai is a general-purpose time series forecasting Transformer based on a masked encoder, pre-trained on the LOTSA dataset as a large-scale time series model.

Moirai is a general-purpose time series forecasting Transformer based on a masked encoder, pre-trained on the LOTSA dataset as a large-scale time series model

A major upgrade of the Moirai-1.0-R model, demonstrating significant improvements across 40 datasets from the Monash repository, with special optimizations for low-frequency data scenarios.

Chronos is a family of pretrained time series forecasting models based on language model architectures, trained by converting time series into token sequences, supporting probabilistic forecasting.

Moirai-1.1-R is a time series forecasting foundation model released by Salesforce, excelling in low-frequency data (such as annual and quarterly data) with approximately 20% performance improvement over its predecessor.

Moirai-1.1-R is a time series forecasting foundation model launched by Salesforce, with significant performance improvements for low-frequency data prediction.

TimeMoE-200M is a billion-scale time series foundation model based on the Mixture of Experts (MoE) architecture, focusing on time series forecasting tasks.

A time series forecasting model fine-tuned based on the amazon/chronos-bolt-base architecture, specifically optimized for intermittent demand forecasting tasks

Timer is an 84-million-parameter lightweight generative Transformer model specifically designed for time series forecasting tasks, supporting zero-shot point prediction.

Intermittent demand forecasting model fine-tuned on 190 million time series data points

A time series forecasting model fine-tuned on proprietary datasets, particularly suitable for intermittent demand forecasting scenarios.

PatchTST is a pre-trained time series foundation model, focusing on time series forecasting tasks.

PatchTSMixer is a time series forecasting foundational model under IBM's Granite project, featuring an innovative Patch hybrid architecture suitable for various time series forecasting tasks.

This is a time series forecasting model based on Transformer architecture, trained for 30 epochs on the tourism-monthly dataset.

A pretrained time series foundation model developed by Google Research, focusing on univariate time series forecasting

YingLong is a foundation model for time series forecasting, pre - trained on 78B time points, providing strong support for time series forecasting tasks.

Lightweight and fast multivariate time series forecasting model, achieving a test set MSE of 0.37 on the ETTh1 dataset

PatchTST is a Transformer-based time series forecasting model designed for long-term time series forecasting, utilizing subsequence patching and channel independence techniques to improve prediction accuracy.

TinyTimeMixer (TTM) is a lightweight foundational model for time series forecasting, focused on efficient time series analysis tasks.

A time series forecasting model fine-tuned based on amazon/chronos-bolt-small, trained with 25 million rows of proprietary data