Method●Since v1.4

invoke

Classify one request.

The run is traced as an LLM run with ls_provider and ls_model_name metadata, and its output carries usage_metadata when the provider reports usage.

invoke<
  Q extends Record<string, __type | __type | __type | __type>
>(
  input: ClassifierRequest<Q>,
  options: Partial<RunnableConfig<Record<string, any>>>
): Promise<ClassifierResponse<Q>>

Parameters

Name	Type	Description
`input`*	`ClassifierRequest<Q>`	State and questions to evaluate.
`options`	`Partial<RunnableConfig<Record<string, any>>>`	Callbacks, tags, metadata, and other runnable config.

View source on GitHub

invoke

Parameters

Chat LangChain

What can I help with?

Menu

invoke

Parameters