ResearchWednesday, September 9, 2026· 2 min read

IBM Releases State-of-the-Art Granite Time Series Model for Businesses

TL;DR

IBM has released Granite Time Series PatchTST-FM-r2, a state-of-the-art foundation model for time series forecasting, on Hugging Face. With a commercial-friendly license, the model can help teams build practical forecasting tools for areas like operations, finance, energy, retail, and infrastructure planning.

Key Takeaways

  • 1IBM’s new Granite Time Series PatchTST-FM-r2 model advances AI-powered forecasting for sequential data.
  • 2The model is available on Hugging Face, making it easier for developers and researchers to access and test.
  • 3A commercial-friendly license supports real-world adoption by businesses and product teams.
  • 4Better time series forecasting can improve planning, reduce waste, and support smarter decision-making across industries.

IBM has released Granite Time Series PatchTST-FM-r2, a state-of-the-art model designed for time series forecasting, through the Hugging Face ecosystem. Time series AI helps analyze data that changes over time, such as demand, traffic, energy use, prices, sensor readings, and operational metrics.

The biggest win is accessibility: by releasing the model with a commercial-friendly license, IBM is making advanced forecasting technology easier for companies, startups, and developers to adopt. That can shorten the path from research to real-world tools that help organizations plan more accurately.

Why it matters

  • Forecasting is everywhere: Businesses rely on predictions to manage inventory, staffing, logistics, maintenance, and risk.
  • Open access accelerates innovation: Hosting the model on Hugging Face makes experimentation and deployment more approachable.
  • Commercial use is encouraged: A business-friendly license can help more teams move from prototypes to production.

This release is a strong example of AI progress becoming more practical and usable. By bringing high-performing time series modeling to a broader community, IBM is helping developers build forecasting systems that can improve efficiency, reduce costs, and support better decisions at scale.

Get AI Wins in Your Inbox

The best positive AI stories delivered to your inbox. No spam, unsubscribe anytime.