Tool use refers to the ability of an AI model to use external tools (APIs, calculators, browsers, databases) to augment its capabilities beyond simple text generation.
What is tool use?
Tool use is the broader concept that encompasses function calling. It refers to an AI model's ability to recognize its limitations and use external tools to overcome them: calculate, search the web, call APIs, manipulate files, etc.
Common tool types
- REST APIs: Retrieve or modify data on external services
- Web search: Access real-time information
- Calculator: Perform precise calculations
- Code execution: Run Python, JavaScript scripts, etc.
- Databases: Query and modify structured data
Tool use vs Function calling
Function calling is the technical mechanism (how the LLM generates the call), while tool use is the concept (the LLM uses tools). Function calling is an implementation of tool use.
Tool use and MCP
The MCP protocol standardizes tool use by defining a universal format for declaring and calling tools. AILabsAudit exposes its features via MCP, allowing any compatible LLM to access audit data, launch analyses and generate reports.