What is Function / tool calling?

A model's ability to call external tools or APIs in a structured way, letting it fetch data or perform actions reliably.

Tool calling (also called function calling) lets a model request that external code be run on its behalf. Developers describe available tools to the model in a structured schema, for example "get_weather(city)" or "search_orders(email)". When the model decides a tool would help, it outputs a structured request (usually JSON) naming the tool and its arguments instead of plain text. The application executes the real function, returns the result to the model, and the model continues with that information. This is how chatbots check live data, book meetings, query databases, or trigger actions, things a text-only model cannot do from training data alone. Practical nuance: the model only proposes the call; your code runs it, so you control permissions and validation. Reliability of picking the right tool with correct arguments varies noticeably between models.

Example

A user asks a travel bot "what is the weather in Denver this weekend?" The model emits a get_weather call with city set to Denver, the app fetches the live forecast from a weather API, and the model writes the answer using real data.

Why it matters

Tool calling is what turns a chatbot into something that acts on live data and real systems. When choosing a model for automation or agents, its tool-calling accuracy matters as much as its writing quality. Browse the AI tools directory or the model leaderboard to put it into practice.