Chinese video AI models beat US rivals in visual quality, but lag on safeguards: HKU study
ByteDance’s Seedance 2.0 Pro ranks first in a University of Hong Kong study, outperforming Google DeepMind's Veo 3.1 and OpenAI's Sora 2 Pro.
Quick Look
Chinese AI video-generation models, led by ByteDance's Seedance 2.0 Pro, have outperformed top US rivals including Google's Veo 3.1 and OpenAI's Sora 2 Pro in visual quality, according to a University of Hong Kong study.
AI-generated summary
Why It Matters
Researchers at the University of Hong Kong evaluated 12 leading AI video-generation models.
Chinese AI video-generation models have surpassed leading US rivals in visual quality, but several fall behind in safety safeguards, according to researchers at the University of Hong Kong.
ByteDance’s Seedance 2.0 Pro took top spot in a ranking of 12 leading artificial intelligence video-generation models, scoring 3.84 out of 5 for visual quality. It outperformed Google DeepMind’s Veo 3.1, which scored 3.47, and OpenAI’s Sora 2 Pro, on 3.27, in a study led by Zhenhui Jack Jiang, professor of innovation and information management at HKU’s business school.
The ByteDance model also led in visual consistency, physical realism, aesthetic quality and instruction compliance, and maintained its lead when audio capabilities were factored in, scoring 3.86 against Veo’s 3.51 and Sora’s 3.35.
Open Questions
- Which specific safety safeguards did the Chinese models fall behind in?
- How large was the sample size and exact methodology of the HKU study?
