Stable Virtual Camera

Project Page

Stable Virtual Camera is a 1.3B generalist diffusion model for Novel View Synthesis (NVS), generating 3D consistent novel views of a scene, given any number of input views and target cameras. We encourage you to visit our website for more visualizations.

Updates

June 2025: Release v1.1 checkpoint fixing known issues of foreground objects sometimes being detached from the background in v1.0.

Model Description

Developed by: Stability AI
Model type: Transformer image-to-video model
Model details: This model was trained to generate 3D consistent novel views of a scene given any number of input views and target cameras. Users can specify the target camera trajectory freely, spanning across a large spatial range. Our model is capable of generating large viewpoint changes and temporally smooth samples. As a result, our samples maintain high consistency without requiring additional NeRF distillation, streamlining the view synthesis pipeline in the wild. Furthermore, we show that our method can generate high-quality videos lasting up to half a minute with seamless loop closure.

License

Non-Commercial License: Free for research and non-commercial use by organizations and individuals. Please refer to Stability AI's Non-Commercial License, available here, for more information.

Model Sources

Project Page: https://stable-virtual-camera.github.io/
Paper: https://arxiv.org/abs/2503.14489
Blog: https://stability.ai/news/introducing-stable-virtual-camera-multi-view-video-generation-with-3d-camera-control
Code: https://github.com/Stability-AI/stable-virtual-camera
Demo: https://huggingface.co/spaces/stabilityai/stable-virtual-camera
Video: https://www.youtube.com/channel/UCLLlVDcS7nNenT_zzO3OPxQ

Usage

For usage instructions, please refer to our GitHub repository.

Intended Uses

Intended uses include the following research and non-commercial uses:

Generation of artworks and use in design and other artistic processes.
Applications in educational or creative tools.
Research on reconstruction models, including understanding the limitations of these models.

All uses of the model should be in accordance with our Acceptable Use Policy.

Out-of-Scope Uses

The model was not trained to be factual or true representations of people or events. As such, using the model to generate such content is out-of-scope of the abilities of this model.

Contact

Please report any issues with the model or contact us:

Security issues: [email protected]
Privacy issues: [email protected]