LTX-Video
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Video Generation | LTX-Video | Aesthetic Score59.3 | 7 | |
| Robustness to purification defenses | LTX-Video (test) | Subject Preservation4.46 | 6 | |
| Text-to-Video Generation | LTX-Video-2B 4x native length (481 frames) (dev) | Subject Consistency0.9726 | 6 | |
| AI-generated video detection | LTX-Video (test) | Accuracy75.33 | 4 |
Showing 4 of 4 rows