On March 6, 2025, Tavus, a leading AI research company backed by Sequoia, introduced three new AI models—Phoenix-3, Raven-0, and Sparrow-0—marking a major leap in human-AI interaction. These models power the next version of Tavus’ Conversational Video Interface (CVI), enabling emotionally intelligent, hyper-realistic video agents.
🔥 Phoenix-3 – The first full-face AI rendering model using Gaussian-diffusion animation.
🦅 Raven-0 – The first AI perception model that sees and reasons like a human.
🐦 Sparrow-0 – A state-of-the-art conversational turn-taking model that eliminates robotic dialogue timing.
These models work together to power Tavus’ Conversational Video Interface (CVI), allowing developers to create video agents that feel truly human.
Demo Persona: "Charlie"
Users on X say Charlie’s ability to express emotion and respond in real time moves AI past the "uncanny valley" into natural interaction.
🔹 Companies like CVS, Alibaba, and Deloitte, along with startups like Delphi and Mercor, are already leveraging Tavus’ AI models.
🔹 Charlie’s demo is now live on the Tavus homepage.
🔹 Developers can start building with these models today.
📢 “Next-level AI”
📢 “A complete OS for human-AI interaction”
📢 “This changes everything about how we talk to AI”
Tavus’ CEO, Hassaan Raza, says the goal is to merge emotional intelligence (EQ) with the intelligence & scalability of AI—and based on early reactions, they’re delivering on that vision.