Home Granite
IBM Granite

IBM® Granite™ AI foundation models are cost-efficient and enterprise-grade. Open-sourced for innovation, they're tailored for business and optimized for scale.

Start building with Granite See Granite documentation
Open and transparent

The future of AI is open. That’s why we strive to make AI as accessible for as many developers as possible. Our open-sourced family of core Granite models is available under an Apache 2.0 license for broad, unencumbered commercial usage, along with tools to monitor the model data—ensuring it’s up to the standards demanded by enterprise applications.

Models Granite for code

Granite decoder-only models are designed for code generative tasks, trained with code written in 116 programming languages.

Explore the IBM Granite Playground Get code models on Hugging Face
Granite for time series

Granite time series models are lightweight and pre-trained for time-series forecasting, optimized to run efficiently across a range of hardware configurations.

Use Granite time series models Get time series models on Hugging Face
Granite for geospatial data

NASA and IBM teamed up to create an AI Foundation Model for Earth Observations using large-scale satellite and remote sensing data.

Try a geospatial model on Hugging Face
Build with Granite

When you’re ready to deploy open-source Granite models in production, Red Hat Enterprise Linux AI and watsonx provide the support and tooling you need to confidently deploy AI in your business at scale.

Explore the tutorials
Create a LangChain AI agent in Python using watsonx

Discover how to build an AI agent that can answer questions

See the tutorial
Create a LangChain RAG system in Python with watsonx

For an LLM to answer questions, fetch the data to create a vector store as context 

See the tutorial
Use foundation models for time series forecasting

Forecast the future based on learning with the TinyTimeMixer (TTM) Granite Model

See the tutorial
Generate SQL from text with LLMs

Convert text into a structured representation and generate a semantically correct SQL query

See the tutorial
Evaluate RAG pipeline using Ragas in Python with watsonx

Use the Ragas framework for Retrieval-Augmented Generation (RAG) evaluation in Python using LangChain

See the tutorial
Build a local AI co-pilot using IBM Granite code, Ollama and Continue

Learn how to adopt AI co-pilot tools in an enterprise setting with open source software

See the tutorial
Prompt tune a Granite model in Python using watsonx

Prompt tune a Granite model in Python using a synthetic dataset containing positive and negative customer reviews.

See the tutorial

IBM TechXchange Conference 2024 | 21–24 October in Las Vegas

See the developer's guide to the conference, and join the must-attend event for technologists using IBM products and solutions.

Register for TechXchange
Granite in the news Leading in open-source LLMs on API calling

IBM’s Granite 20B model tops several benchmarks ranking large language models by how reliably they connect to external software tools.

Best-in-class performance

In testing against a range of other models, including open-source and proprietary, we found Granite models are competitive at a range of coding tasks.

Ranked top 5 in Stanford Transparency Index

A new report from Stanford University’s Center for Research on Foundation Models showed that IBM’s model scored a perfect 100% in several categories designed to measure how open models really are.

Named "strong performer" in Forrester Wave

According to Forrester, the Granite family of models provides enterprise users with some of the most robust and clear insights into the underlying training data.

See the latest updates from IBM Research on our Granite model work

Next steps

Start building with Granite
See granite documentation
Read the IBM statement on IP protection

IBM believes in the creation, deployment and utilization of AI models that advance innovation across the enterprise responsibly. IBM watsonx AI and data platform have an end-to-end process for building and testing foundation models and generative AI. For IBM-developed models, we search for and remove duplication, and we employ URL blocklists, filters for objectionable content and document quality, sentence splitting and tokenization techniques, all before model training.

During the data training process, we work to prevent misalignments in the model outputs and use supervised fine-tuning to enable better instruction following so that the model can be used to complete enterprise tasks via prompt engineering. We are continuing to develop the Granite models in several directions, including other modalities, industry-specific content and more data annotations for training, while also deploying regular, ongoing data protection safeguards for IBM developed models. 

Given the rapidly changing generative AI technology landscape, our end-to-end processes are expected to continuously evolve and improve. As a testament to the rigor IBM puts into the development and testing of its foundation models, the company provides its standard contractual intellectual property indemnification for IBM-developed models, similar to those it provides for IBM hardware and software products.

Moreover, contrary to some other providers of large language models and consistent with the IBM standard approach on indemnification, IBM does not require its customers to indemnify IBM for a customer's use of IBM-developed models. Also, consistent with the IBM approach to its indemnification obligation, IBM does not cap its indemnification liability for the IBM-developed models.

The current watsonx models now under these protections include:

(1) Slate family of encoder-only models.

(2) Granite family of a decoder-only model.

Learn more about licensing for Granite models