Vivix-W1
Streaming-native multimodal model for interactive worlds that evolve in real time
Community-submitted · content unverified. This organization has not verified ownership or the information in this profile.
About
Vivix-W1 is a streaming-native multimodal model that generates audiovisual worlds continuously in fine-grained segments, so text, voice, images, video, touch, camera and motion inputs can change what happens next during generation. It supports native audio-video co-generation, multi-shot storytelling and long-horizon consistency, and Vivix reports roughly 30B active parameters with second-level latency. The public API offers asynchronous 5–15 second world-video generation (vivix-w1, vivix-w1-fast and vivix-w1-ultrafast), while the live streaming variant is in private beta.
- Website
- vivix.ai/vivix-w1
- Added via
- web
- Ownership
- unclaimed
Where to go
Product details
- Product kind
- Api