mirascope.llm.call

A decorator for making provider-agnostic LLM API calls with a typed function.

Usage Documentation

Calls

This decorator enables writing provider-agnostic code by wrapping a typed function that can call any supported LLM provider's API. It parses the prompt template of the wrapped function as messages and templates the input arguments into each message's template.

Example:

from mirascope.llm import call


@call(provider="openai", model="gpt-4o-mini")
def recommend_book(genre: str) -> str:
    return f"Recommend a {genre} book"


response = recommend_book("fantasy")
print(response.content)

Parameters:

Name	Type	Description	Default
provider	str	The LLM provider to use (e.g., "openai", "anthropic").	required
model	str	The model to use for the specified provider (e.g., "gpt-4o-mini").	required
stream	bool	Whether to stream the response from the API call.	required
tools	list[BaseTool | Callable]	The tools available for the LLM to use.	required
response_model	BaseModel | BaseType	The response model into which the response should be structured.	required
output_parser	Callable[[CallResponse | ResponseModelT], Any]	A function for parsing the call response whose value will be returned in place of the original call response.	required
json_mode	bool	Whether to use JSON Mode.	required
client	object	An optional custom client to use in place of the default client.	required
call_params	CommonCallParams	Provider-specific parameters to use in the API call.	required

Returns:

Name	Type	Description
decorator	Callable	A decorator that transforms a typed function into a provider-agnostic LLM API call that returns standardized response types regardless of the underlying provider used.