Choose modules

Modules are components that provide specific capabilities to your agent, such as speech recognition, text generation, and voice synthesis. This guide shows you how to configure modules in the TEN Agent playground.

Before you start:

• Complete the Agent quickstart
• Verify the playground is running at localhost:3000

TEN Agent playground supports the following module types:

• Speech-to-text (STT): Converts spoken language to text
• Text-to-speech (TTS): Converts text to spoken language
• Large language model (LLM): Generates text responses based on input prompts
• Voice-to-voice model (V2V): Transforms voice input into voice output
• Tool (TOOL): Provides additional capabilities like weather updates or news retrieval

1. Open the playground at localhost:3000

2. Select a graph type from the dropdown, such as voice_assistant_integrated_stt or voice_assistant_realtime

3. Click the module button to open the module selection panel

4. Choose modules from the available options for your selected graph type

5. Click Save Changes to apply your selection

   The system displays a confirmation message when modules are successfully configured.

Tools extend the capabilities of LLM and V2V modules. You can bind existing tools such as weather check and news update, or create custom tool extensions.

To bind a tool:

1. Open the module selection panel using the button.

2. Locate the LLM or V2V module in the module picker panel

3. Click the button next to the module name to open tool selection

4. Select an available tool

5. Click Save Changes

   The system confirms when the tool is successfully bound to the module.