https://livekit.io logo
Join Slack
Powered by
I plan to use gpt-5.4 as my LLM. and use livekit s...
# ask-ai
b
I plan to use gpt-5.4 as my LLM. and use livekit sdk for the chat agent, no voice at all. can it be done?
✅ 1
t
Thanks for reaching out, Samuel. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.
Thanks — yes, that can be done. You can use GPT-5.4 (
openai/gpt-5.4
) with LiveKit Agents and run a text-only chat agent by disabling audio in the AgentSession/RoomOptions (
audio_input=False
, 
audio_output=False
). Use LiveKit’s text streams (
lk.chat
, 
lk.transcription
) and `sendText()`/text callbacks to send/receive messages and stream responses. If you’d like, tell me: • which SDK you’re using (Python or Node.js), and • whether you want example code for creating the AgentSession with GPT-5.4 and text-only options. Relevant docs: OpenAI LLM models and Text & transcriptions. Sources: OpenAI LLM models | LiveKit Documentation | Text and transcriptions | LiveKit Documentation | Large language models (LLM) overview | LiveKit Documentation, with interactive elements
b
Python
t
Yes — GPT-5.4 is supported and you can run a text-only chat agent with the Python SDK. Below is a minimal example showing how to configure an AgentSession with 
openai/gpt-5.4
and disable audio so all interaction is text-only.
Copy code
from livekit.agents import AgentSession, inference, RoomOptions

# configure LLM to use GPT-5.4
llm = inference.LLM(model="openai/gpt-5.4")

# disable audio input/output for a text-only session
room_options = RoomOptions(audio_input=False, audio_output=False)

session = AgentSession(
    llm=llm,
    room_options=room_options,
    # other options as needed (callbacks, memory, tools, etc.)
)

# example: send/receive text via the session (frontend uses sendText/send message flows)
# session.start() or integrate with your room lifecycle per your app
If you want, I can: • provide a complete runnable example (including auth and handling incoming messages), • show how to stream text responses to the frontend using `lk.chat`/`lk.transcription`, or • add example callbacks for handling messages and generating replies. Which of those would you like next? , with interactive elements
2 Views
Open in Slack
PreviousNext
Powered by