nemotron-voice-agent-deploy

To deploy a real-time voice agent, you configure NVIDIA ASR, TTS, and LLM services with WebRTC or WebSocket transport using Docker. The deployment process requires an NVIDIA API key and environment setup to enable live speech-to-speech capabilities.

You can run the Nemotron Voice Agent on Jetson Thor, an x86 workstation, or Cloud NIMs. The deployment process includes automatic hardware detection and platform selection to configure the environment accordingly across supported devices.

Setting up real-time speech-to-speech requires Docker, an NVIDIA API key, and proper environment configuration. The deployment guide provides steps to configure NVIDIA ASR, TTS, and LLM with your chosen WebRTC or WebSocket transport protocol.

Yes, the Nemotron Voice Agent supports optional multilingual capabilities for live conversations. During environment setup, you can configure multilingual support and select specific models to enable speech-to-speech interactions in various languages.

The best way to provision edge devices for hands-free voice assistants is using Docker to deploy the Nemotron Voice Agent on platforms like Jetson Thor. This enables integrated ASR, TTS, and LLM for real-time conversational automation workflows.