Function Calling

Estimate the extra tokens a single tool round trip adds compared with an inline answer

Give one training example for SFT'ing a single weather tool call.

In MCP, what exactly is a 'tool' and who decides when to call it?

A 40 team org already has OpenAPI specs and three LLM providers. Defend layering MCP on top instead of picking one integration approach.

Your colleague says MCP and function calling do the same thing. Where are they wrong?

What is function calling in modern LLM APIs, and how does it support agent loops?

What is a 'tool' in the context of an LLM agent, and what does it let the agent do?

What does it mean when an LLM 'makes a tool call'?

What data shape does a tool calling fine-tune use, and what commonly goes wrong?

PremiumCompare MCP and vendor…

Which statement correctly describes the relationship between MCP and vendor function calling?

At which architectural layer does MCP sit versus vendor function calling?

Which statements correctly describe the MCP primitive types?

Compare OpenAI function calling, Anthropic tool use, and ReAct free form text as agent tool invocation mechanisms

Match each tool invocation mechanism to its defining characteristic

What problem does MCP solve that function calling alone does not?

What does the Model Context Protocol (MCP) uniquely enable that provider native function calling does not?

How does MCP differ from OpenAI style function calling?

What is the Model Context Protocol (MCP) and what problem does it solve?