Reasoning level

Some Large Language Models support configurable reasoning or “thinking” levels. In these cases Live Hub automatically sets them to the minimal supported level by default. This helps reduce response latency and makes voice interactions feel more natural and fluid.

Use reasoning_level advanced configuration parameter to override this behavior. For example:

{
    "reasoning_level": "medium"
}

Use reasoning_logs advanced configuration parameter to include reasoning logs in the AI Agent logs, for models that support such functionality (e.g. gemini-2.5-flash-native-audio). For example:

{
    "reasoning_logs": true
}