Guides¶

The guides in this section provide practical, task-oriented examples for common use cases. Each guide focuses on a single topic and includes working code that you can copy and adapt directly.

All examples assume you have already installed the package and have a Temporal server and worker running. See Getting Started if you have not done that yet.

Available Guides¶

Text Responses¶

Learn how to get plain-text answers from web-based LLMs. Covers async and sync usage, single system prompts, multiple system prompts with the @agent.system_prompt decorator, and notes on binary content handling.

Structured Output¶

Use Pydantic models as output_type to get validated, typed responses. Covers basic models, nested models, enums, optional fields, and how the JSON extraction pipeline works internally.

Conversations¶

Build multi-turn conversations by passing message_history between calls. Includes a full async two-turn example and details on how messages are formatted for the Temporal workflow.

Comparing Models¶

Run the same prompt against multiple models concurrently using asyncio.gather. Useful for benchmarking response quality or latency across providers.

For a complete reference of all classes, functions, and exceptions, see the API Reference.