Tool-Augmented LLM
LLMs that can invoke external tools (APIs, code execution, search) to extend their capabilities.
Overview
Tool-augmented LLMs are large language models that can recognize when to use external tools and generate appropriate tool calls to extend their capabilities beyond text generation. Tools include web search, code execution, calculators, databases, APIs, and any external service accessible through structured function calls.
Key Details
The model learns to output structured function calls that are executed by the runtime, with results fed back to the model for further reasoning. This pattern overcomes fundamental LLM limitations: calculators for precise math, search for current information, code interpreters for complex computation. Tool use is central to AI agent architectures and is supported by all major LLM providers (OpenAI, Anthropic, Google).