### Answer Based on the documentation, the Speak During Execution and Speak After Execution settings are described as

Prompt

based—meaning the LLM generates the message based on the config and the conversation context. [(1) ]() There's an important clarification: these options control the agent's behavior when the user is not talking during the function call period. If the user interrupts or talks, the agent would respond normally regardless of these settings. A few possibilities for what you're experiencing: 1. **User interaction triggering speech**: If the user says something during execution, the agent may respond independently of the "Speak During Execution" setting. 2. **Potential bug**: There have been community reports of unexpected behavior with these settings. One user reported that even with both "Wait for Result" and "Speak During Execution" off, the node still appeared to wait instead of transitioning immediately. [(2) ]() 3. **LLM-generated content**: Since these settings are prompt-based, the LLM might be generating speech based on conversation context rather than the explicit setting. [(1) ]() I'd recommend: - Checking if there's any user speech happening during the function call - Reviewing your prompt to ensure it doesn't instruct the agent to speak during function execution - If the issue persists, this may be worth reporting to support with your call ID for investigation

​ To learn more, see the sources I considered: [[1] Integrate any system with custom function](https://docs.retellai.com/build/single-multi-prompt/custom-function) [[2] Wait for Result is off and Speak During Execution is off - custom function is delayed.](https://discord.com/channels/1186424376300679209/1460216156777353340) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer This is expected behavior based on the documentation. The Speak During Execution and Speak After Execution options only control the agent's behavior when the user is not talking during the function call period. [(1) ]() Specifically, the docs state: > "If user interrupts or talks, the agent would respond normally regardless of these settings." So even with "Speak During Execution" off and an empty prompt, if the human speaks during function execution, the agent will respond to them—this is by design, not a bug. If you want to prevent the agent from responding to user speech during function execution, you could try enabling Block Interruptions in the Function Node settings. [(2) ]() For Conversation Flow specifically, when

wait for result

is on and

speak during execution

is off, the agent will transition once the function result is ready. However, "if the user interrupts the agent, the transition can also happen once function result is ready and user is done speaking." If this behavior doesn't fit your use case, you may need to adjust your flow design or consider whether blocking interruptions during the function call would provide a better experience.

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

​ To learn more, see the sources I considered: [[1] Integrate any system with custom function](https://docs.retellai.com/build/single-multi-prompt/custom-function) [[2] Function Node Overview](https://docs.retellai.com/build/conversation-flow/function-node) ​