Expand description
LLM trait, request/response types, and caching configuration.
Structs§
- Citation
Metadata - Citation metadata emitted by model providers for source attribution.
- Citation
Source - One citation source with optional offsets.
- Context
Cache Config - Configuration for automatic prompt caching lifecycle management.
- Generate
Content Config - Configuration for LLM content generation.
- LlmRequest
- A request to an LLM provider.
- LlmResponse
- A response from an LLM provider.
- Usage
Metadata - Token usage statistics from an LLM response.
Enums§
- Finish
Reason - Reason the model stopped generating content.
Traits§
- Cache
Capable - Trait for LLM providers that support prompt caching.
- Llm
- The core trait for all LLM providers.
Type Aliases§
- LlmResponse
Stream - A pinned, boxed stream of
LlmResponseresults from a model.