LightX2V is a lightweight inference framework for image and video generation, built to deliver efficient, high-performance synthesis on both NVIDIA and non-NVIDIA hardware.
X2V โ transform any input modality X (text, image, audio, etc.) into Vision output.
The framework unifies multiple state-of-the-art generation pipelines behind a single runtime, covering training-adjacent inference workflows from research prototypes to production deployment.
Supported Tasks
| Task | Name | Description |
|---|---|---|
| T2V | Text-to-Video | Generate video from a text prompt |
| I2V | Image-to-Video | Animate a still image into video |
| V2V | Video-to-Video | Generate video from a video |
| T2I | Text-to-Image | Generate images from text |
| I2I | Image-to-Image | Edit or transform existing images |
| World Model | Interactive Worlds | Real-time, memory-augmented world simulation |
| โฆ | More | FLF2V, S2V, T2AV, animation, super-resolution โฆ |
See the LightX2V repository for the full, up-to-date task list.
Links
| Resource | URL |
|---|---|
| GitHub | github.com/ModelTC/LightX2V |
| Online Service | x2v.light-ai.top/generate |
| Canvas | x2v.light-ai.top/canvas |
| Blog | light-ai.top/LightX2V-BLOG |
| Blog GitHub | github.com/ModelTC/LightX2V-BLOG |