Frequently Asked Questions (FAQ) about TEN Agent

Q: What is TEN Agent? A: TEN Agent is a conversational AI platform powered by TEN, integrating Gemini 2.0, OpenAI, and RTC for real-time, multimodal interactions.

Q: What are the key features of TEN Agent? A: Key features include real-time vision and screenshare detection, ready-to-use extensions (like Weather Check and Web Search), and integration with Gemini Multimodal Live API.

Q: What are the prerequisites for running TEN Agent locally? A: You need Agora App ID and Certificate, OpenAI API key, Deepgram ASR, FishAudio TTS, Docker, and Node.js (v18+).

Q: How do I set up TEN Agent in my local environment? A: Clone the repository, configure the .env file with your API keys, and use Docker Compose to start the development containers. Then, build and run the agent.

Q: How do I use the Gemini Realtime Extension? A: In the playground at localhost:3000, select the voice_assistant_realtime graph, choose the Gemini Realtime module, and enter your Gemini API key.

Q: What is the relationship between TEN Agent and UBOS? A: TEN Agent is designed to work seamlessly with UBOS, a full-stack AI Agent Development Platform, enhancing its capabilities through orchestration, data integration, and customization.

Q: Where can I get support for TEN Agent? A: You can find support on the Discord server, GitHub Discussions, and GitHub Issues. Also check contribution guidelines.

Q: Is TEN Agent open source? A: Yes, TEN Agent is licensed under the Apache 2.0 License.

Q: What are some potential use cases for TEN Agent? A: TEN Agent can be used for customer support, virtual assistants, remote collaboration, education, and healthcare applications.

Q: Where can I find the TEN Agent playground? A: The playground is accessible at localhost:3000 after setting up and running the agent locally.