In the dashboard, open your agent, go to the Agent tab (agent settings), then open Models. Temperature is configured in the Model configuration there. (elevenlabs.io)
Notes and alternatives:
Temperature is part of the LLM’s Model configuration; lower values are more deterministic, higher values more creative. (elevenlabs.io)
You can also set it programmatically:
At runtime via WebSocket using custom_llm_extra_body.temperature. (elevenlabs.io)
In “agents as code” via CLI config at conversation_config.agent.llm.temperature. (elevenlabs.io)
scenic kestrel
lapis wigeon