OpenAI has introduced o4-mini, a smaller reasoning model designed to handle complex tasks while using fewer resources than larger systems. The company says the model can process up to 200,000 tokens of context, allowing it to work with much longer documents, conversations, and codebases in a single session.

A key feature of o4-mini is native tool calling, which lets developers connect the model to external functions through the Chat Completions API. That capability is aimed at tasks such as retrieving information, running calculations, and interacting with software tools without additional wrapper logic.

OpenAI said the model is available immediately through its API. The release appears to target developers looking for a faster, more efficient option for reasoning-heavy workflows, especially where long context windows matter. The company did not frame the launch as a general consumer product, but as a building block for applications that need structured model output and tool integration.

The announcement adds to OpenAI's effort to broaden its model lineup with systems tuned for different performance and cost tradeoffs. For developers, o4-mini offers another option for building AI products that need extended context and reliable tool use in a smaller package.