Tool calling (also "function calling") is when a language model, instead of answering directly, emits a structured request to invoke a named tool with arguments; the host runs it and feeds the result back to the model.
Tool calling is what turns a text model into something that can act. The model is given tool schemas; when a query needs external data or an action, it outputs a structured call (tool name + JSON arguments) rather than prose. The host executes it and returns the result for the model to use.
It is the mechanism beneath both function-calling APIs and MCP — MCP simply standardizes how tools are described and discovered across providers.
Asked "is the CFO of TSLA buying?", the model emits a tool call `insider.company(ticker="TSLA")` instead of guessing from memory.
Every Arkolith MCP tool is a function the model calls — grounding its answer in our real data instead of its training-time recollection.
Arkolith turns this into live, sourced data your agent can query — SEC filings, insider activity, and market data behind one key, every datapoint traceable to its origin.