ai-infra / API Reference

LLM

from ai_infra.llm import LLM

ai_infra.llm

Extends:BaseLLM

Direct model convenience interface (no agent graph). The LLM class provides a simple API for chat-based interactions with language models. Use this when you don't need tool calling. Example - Basic usage:

python

llm = LLM()
    response = llm.chat("What is the capital of France?")
    print(response.content)  # "Paris is the capital of France."

Example - With structured output:

python

from pydantic import BaseModel

    class Answer(BaseModel):
        city: str
        country: str

    llm = LLM()
    result = llm.chat(
        "What is the capital of France?",
        output_schema=Answer,
    )
    print(result.city)  # "Paris"

Example - Streaming tokens:

python

llm = LLM()
    async for token, meta in llm.stream_tokens("Tell me a story"):
        print(token, end="", flush=True)

Constructor

LLM(callbacks: Callbacks | (CallbackManager | None) = None)

Parameter	Type	Default	Description
`callbacks`	`Callbacks \| (CallbackManager \| None)`	None	Callback handler(s) for observing LLM events. Receives events for LLM calls (start, end, error, tokens). Can be a single Callbacks instance or a CallbackManager.

Methods

from pydantic import BaseModel class Answer(BaseModel): city: str country: str llm = LLM() result = llm.chat( "What is the capital of France?", output_schema=Answer, ) print(result.city) # "Paris"

Parameter

Type

Default

Description

callbacks

Callbacks | (CallbackManager | None)

None

Callback handler(s) for observing LLM events. Receives events for LLM calls (start, end, error, tokens). Can be a single Callbacks instance or a CallbackManager.