EnterpriseModelsTools

IBM drops open-weight Granite 4.2 family with built-in agentic capabilities under Apache 2.0

Source: The Decoder · Matthias Bastian

Intel Summary

IBM has released its Granite 4.2 family of open-weight language models in 3B, 8B, and 30B parameter sizes under the permissive Apache 2.0 license. The models are trained on approximately 15 trillion tokens and support context windows of up to 512,000 tokens. According to IBM, the larger variants utilize agentic reinforcement learning techniques designed to natively improve tool selection and code execution capabilities.

Why It Matters

A commercially permissive Apache 2.0 license combined with a 512,000-token context window gives enterprises a viable, self-hostable alternative to proprietary agent APIs. This release expands options for organizations seeking to build sovereign, customizable multi-step autonomous workflows without incurring per-token API costs or third-party data-sharing risks.

Part of an ongoing development

Independent reporting

IBM released Granite 4.2 model family

IBM has introduced its Granite 4.2 family of large language models, targeting enterprise demand for local and on-premises AI deployment. As open-weight enterprise models, Granite 4.2 is designed to allow organizations to run agentic workflows locally, reducing dependency on external cloud APIs while maintaining control over enterprise data and infrastructure costs. Claims are as reported; this summary makes no determination about accuracy or significance.

Confidence
High confidence
Corroboration
Corroborated

More coverage of this development

Organizations & Entities

Topics