from ai_infra.llm import AudioOutputConfiguration for audio output from LLMs. Use this to request audio responses from audio-capable models.
voice: The voice to use (e.g., "alloy", "nova", "shimmer"). format: Audio format for output (default: "mp3").
from ai_infra.llm import LLM
from ai_infra.llm.multimodal import AudioOutput
llm = LLM()
result = llm.chat(
"Tell me a joke",
model_name="gpt-4o-audio-preview",
audio_output=AudioOutput(voice="nova")
)