TimesFM-3

TimesFM-3

Google
Timestral

Timestral

Mistral AI
TimesFM-3 Timestral

Related Products

  • Yeastar P-Series PBX System
    4.8 out of 5 stars 116 Ratings
    Visit Website for Yeastar P-Series PBX System
  • RaimaDB
    4.8 out of 5 stars 12 Ratings
    Visit Website for RaimaDB
  • Kantata
    4.2 out of 5 stars 2,344 Ratings
    Visit Website for Kantata
  • SKU Science
    4.6 out of 5 stars 16 Ratings
    Visit Website for SKU Science
  • InEight
    4.1 out of 5 stars 168 Ratings
    Visit Website for InEight
  • CompanyCam
    4.7 out of 5 stars 5,588 Ratings
    Visit Website for CompanyCam
  • Runn
    4.8 out of 5 stars 37 Ratings
    Visit Website for Runn
  • Martus
    4.7 out of 5 stars 187 Ratings
    Visit Website for Martus
  • Banker's Dashboard
    4.8 out of 5 stars 6 Ratings
    Visit Website for Banker's Dashboard
  • Planview ProjectAdvantage
    4.3 out of 5 stars 121 Ratings
    Visit Website for Planview ProjectAdvantage
TimesFM-3:

About

TimesFM-3 is a state-of-the-art time series foundation model designed for highly accurate multivariate forecasting in a single forward pass. The 330 million parameter model is pre-trained on a real-world and synthetic time-series corpus comprising more than 1 trillion time points, building on the efficiency and zero-shot generalization of earlier TimesFM models. It can jointly predict multiple coevolving time series and capture dependencies that improve accuracy without task-specific fine-tuning. The model supports multiple targets with point and quantile forecasts, past covariates that are known only historically, and past-future dynamic covariates such as planned promotions, holidays, or weather forecasts. TimesFM-3 uses a decoder-only transformer architecture, processes contiguous data in patches of 32 time steps, and applies alternating causal temporal attention and full variate attention to combine patterns across time and related series.
Timestral:

About

Timestral is a family of time-series foundation models and a Python inference package for generating multivariate probabilistic forecasts from historical sequences. The four model sizes (5M, 28M, 123M and 209M parameters) support past-only, known-future and categorical covariates, return quantile forecasts, and accept context windows up to 32,768 time steps. Developers can load pretrained weights from the Hugging Face Hub or from local files and run inference on CPU or GPU, then incorporate forecasts into monitoring, capacity-planning, energy, supply-chain, financial-risk, or maintenance workflows. Timestral supports zero-shot forecasting and can be adapted to domain data through Mistral Forge. The package is intended for data scientists, machine-learning engineers, and teams building forecasting systems; repository code is Apache 2.0, while model weights use a separate Modified MIT license.
TimesFM-3:

Platforms Supported

Windows Not Supported
Mac Not Supported
Linux Not Supported
Cloud Supported
On-Premises Not Supported
iPhone Not Supported
iPad Not Supported
Android Not Supported
Chromebook Not Supported
Timestral:

Platforms Supported

Windows Not Supported
Mac Not Supported
Linux Not Supported
Cloud Not Supported
On-Premises Supported
iPhone Not Supported
iPad Not Supported
Android Not Supported
Chromebook Not Supported
TimesFM-3:

Audience

Data scientists, researchers, and developers wanting to forecast multiple related time series and incorporate historical and known future signals
Timestral:

Audience

Data scientists, machine-learning engineers, forecasting practitioners, and teams developing time-series forecasting systems.
TimesFM-3:

Support

Phone Support Not Supported
24/7 Live Support Not Supported
Online Supported
Timestral:

Support

Phone Support Not Supported
24/7 Live Support Not Supported
Online Supported
TimesFM-3:

API

Offers API Not Supported
Timestral:

API

Offers API Supported
TimesFM-3:

Screenshots and Videos

Timestral:

Screenshots and Videos

No images available
TimesFM-3:

Pricing

No information available.
Free Version Not Supported
Free Trial Not Supported
Timestral:

Pricing

No information available.
Free Version Supported
Free Trial Not Supported
TimesFM-3:

Reviews/Ratings

Overall 0.0 / 5
ease 0.0 / 5
features 0.0 / 5
design 0.0 / 5
support 0.0 / 5

This software hasn't been reviewed yet. Be the first to provide a review:

Review this Software
Timestral:

Reviews/Ratings

Overall 0.0 / 5
ease 0.0 / 5
features 0.0 / 5
design 0.0 / 5
support 0.0 / 5

This software hasn't been reviewed yet. Be the first to provide a review:

Review this Software
TimesFM-3:

Training

Documentation Supported
Webinars Not Supported
Live Online Not Supported
In Person Not Supported
Timestral:

Training

Documentation Supported
Webinars Not Supported
Live Online Not Supported
In Person Not Supported
TimesFM-3:

Company Information

Google
Founded: 1998
United States
research.google/blog/timesfm-3-a-zero-shot-foundation-model-for-multivariate-forecasting/
Timestral:

Company Information

Mistral AI
Founded: 2023
France
github.com/mistralai/timestral
TimesFM-3:

Alternatives

Timestral

Timestral

Mistral AI
Timestral:

Alternatives

TimesFM-3

TimesFM-3

Google
TimesFM-3:
Timestral:
TimesFM-3:
CodeQwen

CodeQwen

Alibaba
Timestral:
TimesFM-3:
Kimi K2

Kimi K2

Moonshot AI
Timestral:
Solar Mini

Solar Mini

Upstage AI
TimesFM-3:
Qwen-7B

Qwen-7B

Alibaba
Timestral:
Mistral NeMo

Mistral NeMo

Mistral AI
TimesFM-3:

Categories

AI Models Supported
Foundation Models Supported
Timestral:

Categories

AI Models Supported
TimesFM-3:

Integrations

Mistral Forge Not Supported
Timestral:

Integrations

Mistral Forge Supported
TimesFM-3:
Claim TimesFM-3 and update features and information
Claim TimesFM-3 and update features and information
Timestral:
Claim Timestral and update features and information
Claim Timestral and update features and information