Research

The first real-time, human-centric interactive model for continuous perception, conversation, and action in an open world.

Unveiling Vivix-W1: Toward a streaming-native multimodal interactive model for narrative intelligence.

A real-time video runtime that combines offline control, live generation, and intelligent direction to create stories that respond, evolve, and stay alive.

Turning video generation from an offline process into a real-time, interruptible interaction channel built for production at scale.

The idea behind Vivix had been brewing for years. On January 1, 2025, we made the call to start building. The conditions were right, and we saw a path forward that was too compelling to ignore.