Vidu 2.0 Now Live on WaveSpeedAI: A New Generation of Controllable Video Generation Models
We’re excited to introduce the full suite of Vidu 2.0 models—now available on WaveSpeedAI. Designed by Shengshu Technology, these models bring a new level of controllability, visual fidelity, and temporal consistency to open-source video generation.
From image-to-video to reference and start-end controlled synthesis, Vidu 2.0 demonstrates what’s possible when precision meets high-performance multimodal modeling.
With WaveSpeedAI, all models are now available via real-time UI or API—bringing SOTA-level video generation to your browser or product.
Overview: What is Vidu 2.0?
Vidu 2.0 is a family of open-source diffusion-based video generation models, trained on large-scale, high-quality datasets with human-aligned tuning. All three variants on WaveSpeedAI offer distinct entry points into controllable video synthesis, supporting both casual creativity and high-demand content creation pipelines.
With WaveSpeedAI, you can use these models through an intuitive interface or integrate them into custom pipelines with our developer-friendly API.
1. Vidu 2.0 - Image to Video
About
Generate smooth, cinematic videos directly from a single image. The model extrapolates plausible motion and scene dynamics, extending a static photo into a temporally rich narrative sequence.
Features
- High-resolution output (up to 1280×720)
- Temporal coherence up to 80 to 160 frames (4s to 8s)
- Artistic and photorealistic flexibility
- Excellent at facial and motion consistency
Use Cases
- Portrait animation
- Historical photo storytelling
- Stylized photo-to-video transitions
- Product showcase generation
🔗 wavespeed.ai/models/vidu/image-to-video-2.0
2. Vidu 2.0 - Reference to Video
About
Generate videos while preserving the identity and style of a reference image. This model is optimized for facial and visual consistency, making it suitable for avatars, influencers, and branded content.
Features
- Identity-locked generation
- Smooth temporal transitions
- Consistent character motion
- Visual style adherence
Use Cases
- Digital influencers & avatars
- Story-driven video characters
- Fashion or cosplay generation
- Personalization in marketing
🔗 wavespeed.ai/models/vidu/reference-to-video-2.0
3. Vidu 2.0 - Start-End to Video
About
Input the start and end frames, and the model interpolates motion in between, generating a plausible, consistent video sequence. A powerful tool for high-level storytelling and scene transitions.
Features
- Bi-frame guided synthesis
- Strong narrative continuity
- Object-aware and human-aware motion interpolation
- Adaptive to camera movement and layout shifts
Use Cases
- Storyboarding and concept animation
- Scene interpolation in long-form content
- Instructional visual sequences
- Film previsualization
🔗 wavespeed.ai/models/vidu/start-end-to-video-2.0
Try Vidu 2.0 Now on WaveSpeedAI
Vidu 2.0 isn’t just another generative model—it’s a creative toolkit designed for control and clarity. Whether you’re animating stills, preserving identities, or designing story beats between two moments, Vidu 2.0 models enable creators to shape motion with confidence.
All three models are fully available and optimized for real-time performance on WaveSpeedAI. Test them interactively or integrate them via our APIs.
Follow us on Twitter, LinkedIn and join our Discord channel to stay updated.
© 2025 WaveSpeedAI. All rights reserved.