WorldScore
Controllability, quality and dynamics of world generation; Static and Dynamic aggregates
world models and roboticsactive
Score, WorldScore paper table, Dynamic aggregate
Definition
- Version
- not in the notes
- Category
- world models and robotics
- Maintainer
- Duan et al., Stanford (ICCV 2025)
- Status
- active
- Home page
- https://arxiv.org/html/2504.00983v2
Sources
- WorldScore: A Unified Evaluation Benchmark for World Generation arxiv.org https://arxiv.org/html/2504.00983v2
Scores at a Glance
Score, WorldScore paper table, Dynamic aggregate
Higher is better for this unit.
1
CogVideoX-I2VZ.ai (Zhipu)
Gen-3Runway
WonderWorldStanford University
LucidDreamerSeoul National University
CogVideoX-I2VZ.ai (Zhipu)59.12 score
maintainer-run2Gen-3Runway57.58 score
maintainer-run3WonderWorldStanford University50.88 score
maintainer-run4LucidDreamerSeoul National University49.28 score
maintainer-run020406080
Score, WorldScore paper table, Static aggregate
Higher is better for this unit.
1WonderWorldStanford UniversityLucidDreamerSeoul National UniversityCogVideoX-I2VZ.ai (Zhipu)Gen-3Runway
WonderWorldStanford University72.69 score
maintainer-run2LucidDreamerSeoul National University70.4 score
maintainer-run3CogVideoX-I2VZ.ai (Zhipu)62.15 score
maintainer-run4Gen-3Runway60.71 score
maintainer-run020406080
Where this comes from
Each bar is one model, with its most independent recorded score. Hatched bars are vendor-reported; the word beside each bar says who ran it. Settings and dates differ between rows; the Scores table below has every result.
Scores
| Model | Score | Who ran it | Setting | View or population | Date | Source |
|---|---|---|---|---|---|---|
| WonderWorld | 72.69 score | maintainer-run | Static aggregate | WorldScore paper table | 2025 | arxiv.org |
| LucidDreamer | 70.4 score | maintainer-run | Static aggregate | WorldScore paper table | 2025 | arxiv.org |
| CogVideoX-I2V | 62.15 score | maintainer-run | Static aggregate | WorldScore paper table | 2025 | arxiv.org |
| Gen-3 | 60.71 score | maintainer-run | Static aggregate | WorldScore paper table | 2025 | arxiv.org |
| CogVideoX-I2V | 59.12 score | maintainer-run | Dynamic aggregate | WorldScore paper table | 2025 | arxiv.org |
| Gen-3 | 57.58 score | maintainer-run | Dynamic aggregate | WorldScore paper table | 2025 | arxiv.org |
| WonderWorld | 50.88 score | maintainer-run | Dynamic aggregate | WorldScore paper table | 2025 | arxiv.org |
| LucidDreamer | 49.28 score | maintainer-run | Dynamic aggregate | WorldScore paper table | 2025 | arxiv.org |
Where this comes from
Scores from different settings, harnesses or dates are not directly comparable; the row says which. Sorted by score, highest first, as a reading aid, not a ranking.
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "WorldScore", https://aurion.semai.id/benchmarks/worldscore/, records of October 2026.