LiveKit Community

when i use generate_reply and i pass some instructions to it does it also uses the current agent instructions?

Thanks for reaching out, Test. I'm an AI assistant and I'm here to help. Give me a moment while I research this topic.


I’d be happy to help you understand how the `generate_reply` function works depending on the model you’re using:
• *Non-Realtime LLM Models:* Here, the function thoughtfully combines the agent’s current instructions with the ones you provide, linking them with a newline character. This way, both sets of instructions work together smoothly.
• *Realtime Models:* In this case, the function sends only the instructions you pass directly to the server. Usually, the agent’s base instructions are set up during session initialization, so they aren’t combined again during the `generate_reply` call.
If anything feels unclear or if you have a particular scenario in mind, please don’t hesitate to reach out—I’m here to help!
 Sources: <https://github.com/livekit/agents/blob/f2fb5705c865ee7f220f60b5baeaf4c716db4cf6/livekit-plugins/livekit-plugins-google/livekit/plugins/google/beta/realtime/realtime_api.py#L447-L492|livekit-plugins/livekit-plugins-google/livekit/plugins/google/beta/realtime...> | <https://github.com/livekit/agents-js/blob/95ff351cad582443d8163d9487377d48842db8f7/plugins/openai/src/realtime/realtime_model.ts#L566-L628|plugins/openai/src/realtime/realtime_model.ts>