TikTok’s parent company, ByteDance, launched “Jimeng AI,” a new application in China that generates short videos from text prompts.

Jimeng AI produces brief video clips, typically under five seconds, based on user-provided text descriptions. 

The app also offers image generation capabilities. While the AI-generated content exhibits a distinctive hyper-real quality, it surpasses static images in terms of animation and lifelike qualities.

ByteDance’s flagship product, TikTok, already incorporates various generative AI features. These include chatbots, search tools, music generation options, and profile picture simulators.

The company is concurrently developing MagicVideo AI creation tools.

Burning Trend in the AI Industry

ByteDance’s development of text-to-video technology aligns with the broader AI industry trend of expanding creative tools. 

OpenAI unveiled its text-to-video model Sora in February, and Chinese companies have massively developed similar tools.

Kuaishou, one of China’s largest video apps, opened its Kling AI text-to-video model to a global audience late last month.

Chinese AI startup Zhipu AI also introduced its video-generating model Ying last month, and days after its launch, another startup, Shengshu officially launched its Vidu app.

Lastly, Alibaba Group Holding is developing a video-generating tool called Tora, based on Sora.

Jimeng AI is currently only available in China. 

Whether ByteDance plans to launch the app internationally or integrate its features directly into TikTok remains unclear, especially considering the U.S. government is pressuring the company to divest its U.S. TikTok operations.