Structured Output
The ability of an LLM to generate responses in a specific format like JSON, XML, or a defined schema. Structured output makes AI responses parseable by other software systems.
Why It Matters
Structured output is essential for integrating LLMs into software applications. Without it, parsing free-text responses is unreliable and error-prone.
Example
An LLM extracting entities from text and returning: {"name": "John Smith", "company": "Acme Corp", "role": "CEO", "sentiment": "positive"} instead of free text.
Think of it like...
Like filling out a form versus writing a letter — the form ensures information is in specific, predictable slots that can be automatically processed.
Related Terms
Function Calling
A capability where an LLM can generate structured output to invoke specific functions or APIs. The model decides which function to call and what parameters to pass based on the user's request.
API
Application Programming Interface — a set of rules and protocols that allow different software applications to communicate with each other. In AI, APIs let developers integrate AI capabilities into their applications.
Tool Use
The ability of an AI model to interact with external tools, APIs, and systems to accomplish tasks beyond text generation. Tools extend the model's capabilities to include search, calculation, code execution, and more.