China’s ‘AI Tiger’ Zhipu in race to bring enhanced features to video-generation apps

China’s state-backed unicorn Zhipu AI has introduced a series of artificial intelligence (AI) features to its video tool Ying, in a heated race with other start-ups and Big Tech companies to woo users in a growing market for AI-assisted video creation tools.

After the updates, Ying will be able to generate 4K high-definition video clips as long as 10 seconds at 60 frames per second, Zhipu said in a statement. The Beijing-based firm is considered one of China’s four “AI Tigers”, a group of unicorns representing the mainland’s best hope in narrowing its technological gap with the US.

Ying was released to the public in July after leading US generative AI start-up OpenAI unveiled in February its text-to-video model Sora. OpenAI’s services are officially unavailable in China. Since then, a bevy of Chinese tech giants and start-ups have launched a string of AI-powered video creation tools.

TikTok parent ByteDance announced on Friday an update to its Jimeng AI video app that includes the second iteration of its Seaweed video model to enable faster generation of video clips with richer details.

ByteDance is also adding to its Doubao chatbot a new video function for converting text or images into lifelike video clips, thanks to its “outstanding semantic understanding capability”, it said.