Technology
Tools parameter
The `tools` parameter is the array that registers external functions and capabilities with a Large Language Model (LLM), enabling it to generate structured JSON for real-world actions.
This critical API parameter, typically an array of objects, injects a JSON schema of available functions into the LLM’s context window: it’s how you define the model’s external capabilities. Specifically, the model analyzes the user's prompt against the provided tool definitions (name, description, and parameter schema) to determine if an external action is required. If so, it returns a `tool_calls` object containing the function name and arguments, such as `get_weather(location='Tokyo')`. Your application then executes this external function—like querying a database or calling a third-party API—and passes the result back to the model for a final, grounded response.
What builders pair with Tools parameter
Projects using both technologies. Select a pairing to see a project.
Pairing: GPT
My first full-stack LLM app: a demo and learnings
Pairing: LangChain
My first full-stack LLM app: a demo and learnings
Pairing: OpenAI API
My first full-stack LLM app: a demo and learnings
Pairing: RAG
My first full-stack LLM app: a demo and learnings
Recent Talks & Demos
Showing 1-1 of 1