Stories about Granite 4.2

IBM has released the Granite 4.2 series of language models, which come in three sizes: 3B, 8B, and 30B, trained on approximately 1.5 trillion tokens, with a maximum context window of up to 512,000 tokens. The larger models were trained using "proxy RL" and are able to learn tool use and code execution autonomously. All models are released under the Apache 2.0 license.

IBM released Granite 4.2, an open family of reasoning language models in three sizes: 3B, 8B, and 30B, all under the Apache 2.0 license. All models support native tool calling and thinking/low-effort/non-thinking switching. The 8B and 30B models were also trained with agent RL and can edit code, drive terminals and run web searches in a sandbox environment.