Vidu S2: Real-Time Interactive, Editable, and Spatial Video Generation
Abstract
Vidu S2 introduces real-time interactive avatar and video editing models that support high-resolution spatial video generation and dynamic reference updates.
We present Vidu S2, which comprises Vidu S2-Avatar, a real-time interactive digital-character model, and Vidu S2-Editing, a real-time video editing model. Moreover, we explore the feasibility of real-time spatial video generation for both Vidu S2-Avatar and Vidu S2-Editing. Compared with Vidu S1, Vidu S2-Avatar supports real-time 720p video generation, generation with dynamic references that can be updated at any moment, and stronger instruction following, such as dancing. Vidu S2-Editing supports editing a video stream in real time, including style rendering, clothing replacement, character replacement, and background replacement. Experiments show that Vidu S2 outperforms all baselines. A playable online demo is available at https://vidu.com/vidu-stream.
Community
🚀 Thrilled to introduce Vidu S2: real-time AI video you can talk to, direct, and real-time editing.
🎙️ Vidu S2-Avatar
Interactive characters at 720p and 25–42 FPS, with stronger instruction following, expressive full-body motion, and dancing. Introduce new reference images anytime to change outfits, interact with objects, or switch scenes mid-conversation.
🎨 Vidu S2-Editing
Transform an incoming video stream in real time while preserving motion and timing: style transfer, virtual try-on, character replacement, and background replacement.
🥽 Spatial video
Bring generated characters and live video edits into stereoscopic video for immersive VR experiences.
🏆 SOTA results across all five public benchmarks in our evaluation, spanning avatar generation and video editing.
🎮 Try it: https://vidu.com/vidu-stream
🛠️ API: http://platform.vidu.com/vidu-stream/doc
📄 Technical report: https://arxiv.org/pdf/2609.11638
Get this paper in your agent:
hf papers read 2609.11638 Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash Models citing this paper 0
No model linking this paper
Datasets citing this paper 0
No dataset linking this paper
Spaces citing this paper 0
No Space linking this paper