Open. Performant. Trusted. Apache 2.0 licensed. Cryptographically signed1. ISO certified2.
Lightweight, performant models, released under an Apache 2.0 license, designed for scalable, enterprise workloads
Why build with Granite?
Build and scale AI faster with customizable, open-source models optimized for enterprise workloads, cost efficiency and flexible deployments.
Open
Open source under Apache 2.0, Granite ensures transparency, while enabling full customizability and deployment flexibility across any infrastructure.
Performant
The small, high-performing models are designed to maximize efficiency and scalability for essential enterprise tasks
Trusted
Eliminate the risk of “black box” AI with transparency into training data and processes, harm detection capabilities and built-in guardrails.
Our most performant dense, non-thinking models yet. Competitive with larger, thinking models across a range of enterprise tasks—at a fraction of the cost.
Small yet powerful. Industry-leading transcription accuracy across accents, domains and noisy environments.
Understand documents, charts and images with enterprise-grade precision.
Guardrails to detect malicious content and harmful outputs. Built for enterprise compliance.
Accurate semantic representations for retrieval, search and classification.
Explore the benchmarks
These models were evaluated against a large collection of datasets and metrics to cover different aspects of text generation. See additional benchmarks in the Granite technical blog.
Access and build
Performance and efficiency
Granite 4.1 delivers competitive instruction‑following and tool‑calling performance without relying on long chains of thought, offering predictable latency, stable token usage and lower operational cost. This makes it a strong, production‑ready choice for enterprise workloads where efficiency and reliability matter most.
- Toll calling
- Instruction following
- Harm detection
- Transcription accuracy
- Table and chart extraction
Granite 4.1 language models understand and execute tool-based instructions, enabling seamless integration with various software tools and APIs. This capability allows enterprises to create powerful AI-driven workflows while automating complex tasks.
Granite 4.1 language models comprehend and adhere to user instructions, ensuring reliable and accurate task completion. This capability is particularly valuable for enterprises looking to automate processes and provide consistent, high-quality results.
Granite Guardian 4.1 detects key risk dimensions catalogued in the IBM AI Risk Atlas. Trained on unique data comprising human annotations and synthetic data from internal red-teaming, Guardian outperforms similar models on standard benchmarks (including, but not limited to, jailbreak attempts, profanity and hallucinations related to tool calls and retrieval augmented generation in agent-based systems).
Granite Speech 4.1 delivers highly accurate, enterprise-ready speech recognition across diverse real-world audio environments, achieving low word error rates on benchmarks spanning conversational speech, meetings, presentations and earnings calls.
Granite Vision 4.1 delivers industry-leading performance in extracting structured information from visual content, achieving the highest average score across seven benchmarks spanning chart extraction, table extraction and key-value pair (KVP) extraction.
Trusted by companies across all industries
US Open
US Open wanted to engage global fans with ever-evolving digital experiences. IBM helped transform massive match data into AI-driven insights and interactive features, delivering a dynamic app and website experience that keeps fans connected and immersed in every moment.
14M
million fans around the globe entering world-class digital experiences
7M
data points captured and analyzed throughout the tournament
Granite for developers
Build a document summarizer with IBM Granite to process documents beyond context window limits.
Build a RAG pipeline with Granite to answer queries using an external knowledge base.
Build a multimodal RAG pipeline with Granite and Docling to query text, tables, and images.
See how open-source LLMs enable autonomy, cut costs, and help developers with evaluation, tuning, and deployment.
Use Granite time series models to perform zero-shot and fine-tuned time series forecasting.
Granite recipes for agentic tasks.
Build a local AI co-pilot using IBM Granite Code, Ollama, and Continue.
View the full Granite Cookbook
Build with Granite
Granite models drive the AI behind many IBM products and services. Discover ready-to-use solutions for code generation, application development, and model testing. All powered by IBM Granite.
Speed up coding and streamline development with AI and automation leveraging Granite models.
Build and deploy AI applications using Granite models or select from a variety of third-party models.
Develop and manage AI agents powered by Granite and explore the catalogue of pre-built agents.
Develop, test and run LLMs, including Granite.
Analysts and leaderboards
Read the report to see how IBM empowers data scientists and machine learning engineers to build, deploy, and govern impactful AI applications across their enterprises.
Discover why Granite Guardian models rank in six of the top 10 spots on the GuardBench Leaderboard, excelling at identifying harmful or malicious prompts and LLM-generated responses.
Explore the leaderboard in which IBM is ranked as the top speech models based on low word error rate.
Explore why Granite earned a top rating in the Stanford Foundation Model Transparency Index, which assesses openness around data sources, dataset size, harmful content filtering, and other key transparency factors.
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.
1As of 29 April 2026, released Granite language, vision, speech, embedding and guardian models are being cryptographically signed.
2ISO certification is for the Granite AI Management System (AIMS) of the Granite language models. The certificate may be found here: https://www.schellman.com/certificate-directory under certificate no. 1102257-1.