@dudleykong Yes, absolutely! You can control which parts of the output are passed to the Text-to-Speech (TTS) system. To achieve this, you would typically structure your code to separate the intermediate processing steps from the final output. Only the final output, which you want the user to hear, should be sent to the TTS engine. Here's a general approach you can take: 1. **Intermediate Processing**: Run the LLM with the necessary prompts to perform the intermediate steps. Capture the output of these steps in variables within your code, but do not send this output to TTS. 2. **Final Solution**: Once you have the final solution from the LLM, you can then pass only this part of the output to the TTS system. This way, the user will only hear the final solution, and the intermediate steps will remain as part of the backend processing. If you need assistance with implementing this in your system or have any specific questions about the process, feel free to ask!

You can continue this conversation by mentioning me in the message.