Tavus

Tavus is an API-first platform for building lifelike conversational video agents and personalized avatar videos. Its Conversational Video Interface combines speech recognition, language models, text-to-speech, perception, turn-taking, WebRTC, memory, knowledge retrieval, function calling, and low-latency video rendering. Developers can choose stock AI humans or train a consented custom Replica from a short video, then embed multilingual face-to-face agents or generate asynchronous avatar videos through APIs, React components, or a no-code portal.