ByteDance, the company behind TikTok, has released a text-to-video AI called Jimeng AI in China. The app, developed by ByteDance subsidiary Faceu Technology, is now available exclusively for Chinese users on iOS and Android.
Jimeng AI comes from ByteDance's Jianying division, best known for the popular video editing app CapCut. Integrating Jimeng into CapCut could significantly boost its reach and relevance.
Jimeng AI-generated videos are automatically formatted for sharing on TikTok, TechRadar notes, though there's no direct connection between the apps. Users can create 26 videos and 80 images for free, with paid plans starting at around $10 per month.
ByteDance has not released technical details about Jimeng AI's architecture or training data.
Initial tests by South China Morning Post show inconsistent results from Jimeng AI. While some inputs produced smooth videos, others led to distorted images and jerky movements.
Compared to OpenAI's Sora, Jimeng AI's output appears less detailed and struggles with complex motions, resembling more of an animated slideshow than cutting-edge AI video.
Video: ByteDance
The launch of Jimeng AI comes at a time of increasing interest in AI video generation, with companies like OpenAI, Runway, and Midjourney developing their own models. Although OpenAI's Sora has attracted significant attention, it is not yet available for public use, leaving room for competitors to enter the market.
Jimeng is not ByteDance's first venture into video AI. The company introduced MagicVideo in 2022, followed by a second-generation version in early 2024. Recently, ByteDance announced plans to go "all-in" on generative AI and substantially expand its workforce in this field. With TikTok and CapCut, the company has a good starting point for selling generative AI services to users.
