Chinese AI start-up Zhipu introduced its latest video generation model, Ying, on Friday, marking a significant step in the growing competition among local tech firms in the AI video space.
The Ying model, available through Zhipu AI’s ChatGLM chatbot, can generate six-second video clips from text and image prompts within approximately 30 seconds. Users can customize these clips using various styles, including 3D animation, cinematic effects, or an oil painting look. Additionally, the model offers emotional themes such as tense, lively, and lonely. The service is available for unlimited use, though free users may experience longer wait times during peak hours.
Comparison of AI Video Models
| Model | Company | Video Length | Customization Options | Accessibility |
|---|---|---|---|---|
| Ying | Zhipu | 6 seconds | 3D animation, cinematic, oil painting, various emotions | Unlimited free use |
| Kling | Kuaishou | Variable | Limited to 6 videos/day for free users | Paid plans available |
| Sora | OpenAI | TBD | Under development | Not yet publicly available |
Zhipu’s launch of Ying comes shortly after Kuaishou’s introduction of the Kling video model, available for limited test use. Kuaishou, a competitor of ByteDance’s Douyin (the Chinese version of TikTok), offers annual paid plans for Kling, allowing for up to 800 videos per month. In contrast, OpenAI’s Sora model, announced in February, remains under development, with the company focusing on preventing misuse of the technology.
Zhipu CEO Zhang Peng noted that Ying’s technology, called CogVideoX, shares similarities with the diffusion transformer (DiT) architecture used by OpenAI’s Sora but boasts faster video generation speeds. Zhang also mentioned that Zhipu is working on an updated version of the video model capable of producing longer, higher-definition videos.
Featured Image courtesy of Tech Times
Follow us for more updates on Zhipu’s latest releases.