Skip to main content
ToolPotion
AI Advisor
Submit a Tool (Free)

Timestral-209M

No ratings yet
AI ModelsForecasting & Predictive AnalyticsOpen source · Web

Timestral-209M is a time series foundation model designed for zero-shot forecasting. It supports univariate and multivariate forecasting, probabilistic predictions, and can handle up to 32,768 historical time steps. The model is open-source, allowing users to reproduce benchmark results easily.

View Model
Share
View Model
Share

Description

Timestral-209M is part of the Timestral family of time series foundation models, specifically designed for zero-shot forecasting. This model is available in a 209M parameter version and supports both univariate and multivariate forecasting. It is capable of handling past-only and known-future covariates, and provides probabilistic predictions across 21 quantiles.

Read More

The model architecture is based on a patch-based transformer with time attention and variate attention, consisting of 24 layers with an embedding dimension of 768. Each patch contains 32 time steps, and the context length can extend up to 32,768 time steps. Timestral-209M can forecast up to 1,024 steps in a single forward pass, with longer horizons forecasted block by block.

Timestral-209M has demonstrated state-of-the-art zero-shot accuracy across major forecasting benchmarks such as GIFT-Eval, TIME, and fev-bench. It outperforms other models in metrics like MASE and CRPS, showcasing its efficiency and reliability in time series forecasting.

The model is open-source, with its code available on GitHub, enabling users to run Timestral-209M and reproduce benchmark results. It operates on PyTorch and can be used on both CPU and GPU, providing flexibility for various computational environments. The model is released under a Modified MIT license, ensuring accessibility while respecting intellectual property rights.

  • CategoryForecasting & Predictive Analytics
  • Pricing ModelOpen source
  • DeploymentCloud
  • PlatformsWeb

Timestral-209M Highlights

  • Zero-shot forecasting

  • Univariate and multivariate support

  • Probabilistic predictions across 21 quantiles

  • Patch-based transformer architecture

  • 24 layers with embedding dimension 768

  • Context length up to 32,768 time steps

  • Forecast horizon up to 1,024 steps

  • Open-source with GitHub repository

Getting Started with Timestral-209M

  1. Access model: Visit Hugging Face

  2. Authenticate: Set up account

  3. Set up environment: Configure PyTorch

  4. Integrate via API: Use model in applications

  5. Optimise: Adjust parameters for specific needs

Timestral-209M's Use Cases

  • Financial Forecasting
  • Supply Chain Management
  • Weather Prediction
  • Energy Consumption
  • Healthcare Analytics

FAQ from Timestral-209M

From Mistral AI

Timestral-209M Reviews

Timestral-209M Reviews

Loading...

Popular AI Tools Like Timestral-209M

Chronos-2 — Time Series Forecasting Model

AI Models

Chronos-2 is a 120M-parameter encoder-only time series foundation model designed for zero-shot forecasting. It supports various forecasting tasks and is efficient, delivering over…

Forecasting & Predictive Analytics
View Tool

Chronos-2 — Time Series Forecasting Model

AI Models

120M-parameter encoder-only time series foundation model designed for zero-shot forecasting. It supports various forecasting tasks and is efficient, delivering over…Chronos-2 is a 120M-parameter encoder-only time series foundation model designed for zero-shot forecasting. It supports various forecasting tasks and is efficient, delivering over…

FeaturedForecasting & Predictive Analytics
See the Top 5 AI Models

TimesFM 3.0 (PyTorch)

AI Models

TimesFM 3.0 is a pretrained time-series foundation model developed by Google Research. It is designed for time-series forecasting and is available with PyTorch weights and…

Forecasting & Predictive Analytics
View Tool

TimesFM 3.0 (PyTorch)

AI Models

Pretrained time-series foundation model developed by Google Research. It is designed for time-series forecasting and is available with PyTorch weights and…TimesFM 3.0 is a pretrained time-series foundation model developed by Google Research. It is designed for time-series forecasting and is available with PyTorch weights and…

Forecasting & Predictive Analytics
See the Top 5 AI Models

PatchTST

AI Models

PatchTST is an official implementation of a Transformer-based model for long-term time series forecasting. It segments time series into patches and processes them independently,…

AI Models & LLMs
View Tool

PatchTST

AI Models

Official implementation of a Transformer-based model for long-term time series forecasting. It segments time series into patches and processes them independently,…PatchTST is an official implementation of a Transformer-based model for long-term time series forecasting. It segments time series into patches and processes them independently,…

AI Models & LLMs
See the Top 5 AI Models

TimesNet

AI Models

TimesNet is a foundation model for general time series analysis, transforming 1D series into 2D space for enhanced modeling. It achieves state-of-the-art results in forecasting,…

AI Models & LLMs
View Tool

TimesNet

AI Models

Foundation model for general time series analysis, transforming 1D series into 2D space for enhanced modeling. It achieves state-of-the-art results in forecasting,…TimesNet is a foundation model for general time series analysis, transforming 1D series into 2D space for enhanced modeling. It achieves state-of-the-art results in forecasting,…

AI Models & LLMs
See the Top 5 AI Models

Prophet

AI Frameworks

Prophet is an open-source forecasting procedure developed by Facebook. It is designed for time series data with strong seasonal effects and multiple seasons. Prophet is fast,…

Forecasting & Predictive Analytics
View Tool

Prophet

AI Frameworks

Open-source forecasting procedure developed by Facebook. It is designed for time series data with strong seasonal effects and multiple seasons. Prophet is fast,…Prophet is an open-source forecasting procedure developed by Facebook. It is designed for time series data with strong seasonal effects and multiple seasons. Prophet is fast,…

Forecasting & Predictive Analytics
See the Top 50 AI Frameworks

Informer2020

AI GitHub Repos

The Informer2020 GitHub repository provides the PyTorch implementation for the Informer model, designed for efficient long sequence time-series forecasting. It features ProbSparse…

AI Models & LLMs
View Tool

Informer2020

AI GitHub Repos

The Informer2020 GitHub repository provides the PyTorch implementation for the Informer model, designed for efficient long sequence time-series forecasting. It features ProbSparse…

AI Models & LLMs
See the Top 250 AI GitHub Repos

NVIDIA Nemotron 3 Ultra

AI Models

NVIDIA Nemotron 3 Ultra is a powerful AI model designed for complex reasoning and multilingual tasks. With 550 billion parameters, it excels in long-context analysis and tool use,…

AI Models & LLMs
View Tool

NVIDIA Nemotron 3 Ultra

AI Models

Powerful AI model designed for complex reasoning and multilingual tasks. With 550 billion parameters, it excels in long-context analysis and tool use,…NVIDIA Nemotron 3 Ultra is a powerful AI model designed for complex reasoning and multilingual tasks. With 550 billion parameters, it excels in long-context analysis and tool use,…

FeaturedAI Models & LLMs
See the Top 5 AI Models

GluonTS

AI Frameworks

GluonTS is a Python package for probabilistic time series modeling, leveraging deep learning models built on PyTorch and MXNet. It facilitates accurate forecasting by generating…

AI Models & LLMs
View Tool

GluonTS

AI Frameworks

Python package for probabilistic time series modeling, leveraging deep learning models built on PyTorch and MXNet. It facilitates accurate forecasting by generating…GluonTS is a Python package for probabilistic time series modeling, leveraging deep learning models built on PyTorch and MXNet. It facilitates accurate forecasting by generating…

AI Models & LLMs
See the Top 50 AI Frameworks
1 of 8