IBM has released Granite 4.2, a family of dense, decoder-only reasoning large language models in 3B, 8B, and 30B sizes. These models are pre-trained from scratch on approximately 15 trillion tokens and feature a context window extended to 512K tokens.
- The training pipeline includes supervised fine-tuning on chain-of-thought and agentic-trajectory data, followed by multi-stage reinforcement learning.
- The 8B and 30B models undergo agentic RL to learn tool use within real sandboxed environments, including software engineering and web search.
- All models support native tool calling via an OpenAI-compatible endpoint and include a switch for thinking or non-thinking modes.
- The models are released under the Apache 2.0 license.