Tool Use

Tool Use is one of the key design patterns within Agentic Reasoning. It refers to an LLM calling external tools, like a search engine, a calculator, a code interpreter or an API, to do things it can't do with text generation alone.

Typically, the model outputs a structured tool call, the system runs the tool, and the result is fed back into the model's context so it can continue. ReAct: Synergizing Reasoning and Acting in Language Models is an early example of interleaving reasoning with actions like this (Yao et al., 2023). MCP is a protocol for connecting models to tools and data sources.

Tool use is the clearest point where LLMs become AI Agents, since they can take actions in the world rather than just returning text.