Changing the LLM (AI Model)
Published:
Overview
This guide explains how to change the Large Language Model (LLM) used by an AI Voice Agent. The LLM is the core AI model that powers the agent’s understanding and responses. Different models offer different capabilities, and the choice of model also affects which voices are available.
Steps to Change the LLM
- Navigate to Call Flows and locate the agent you want to edit.
- Ensure you are in the Advanced View.
- Locate the LLM selector within the agent settings.
- Choose from the available models:
- OpenAI models (e.g., GPT series)
- Google models (e.g., Gemini)
- Open source models
- Speech-to-speech models
- Once selected, click Save to Draft or Publish to Production.
Voice Compatibility by Model Type
| Model Type | Voice Options |
|---|---|
| Text-to-speech models (e.g., OpenAI, Gemini) | Full range of available platform voices |
| Speech-to-speech – Grok | Limited to Grok-specific voices only |
| Speech-to-speech – Gemini | Limited to Google Gemini voices only |
Key Notes
| Item | Details |
|---|---|
| LLM Definition | The LLM is the AI model that powers the agent’s reasoning and responses |
| Voice Impact | Switching to a speech-to-speech model changes the available voice options |
| Speech-to-Speech | Offers more natural, lower-latency responses but has a more limited voice selection |
| Advanced View | Must be enabled to access the LLM setting |
