VGI-Bench: Probing Visual Intelligence in Video Generation Models
Recent studies suggest that video generation models can exhibit certain forms of zero-shot visual reasoning through generated frames. Yet reliable evaluation remains challenging: benchmarks should adopt inputs aligned with the visual priors of current video models, require valid evolving processes rather than only plausible final states, and calibrate task difficulty to remain challenging yet partly feasible. To this end, we introduce VGI-bench, containing 27 tasks and 810 instances, organized by a two-level taxonomy of task domains and skill tags for fine-grained evaluation of visual reasoning capabilities of video generation models. Our evaluations show that current generative systems c...
| Authors | Xuan He, Cong Wei, Yuhao Cheng et al. |
|---|---|
| Primary category | cs.CV |
| Published | 20 August 2026 |
| Last updated at source | 26 August 2026 |
Models and datasets referencing this paper
- hexuan21/VGI-BenchDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.