TikTok's parent company launches AI text-to-video generation tool

ByteDance, known in the US for creating the wildly popular (and controversial) short-form video TikTok app, has officially entered the AI space with its latest release — Jimeng AI. 

As first spotted by Reuters on Tuesday, ByteDance-owned Faceu Technology made Jimeng AI, a text-to-image and text-to-video generator, available on the Apple App Store for Chinese users. According to the report, this follows the app’s Android release on July 31. 

Also: Reddit tests AI-powered search results, mulls paid subreddits

Users can use simple text prompts to generate images and videos within the app. The examples provided on the Jimeng AI website show impressive renditions with realistic details and high-quality renderings. However, the video clips are short, lasting only a few seconds. 

Despite being free to download, the app offers several subscription tiers, including a monthly subscription of ¥69 ($9.65) and a yearly subscription of ¥659 ($91.77), as seen by its Apple App Store listing. With the plans, users can generate roughly 2,050 images or 168 videos monthly, according to the report.

One of the app’s biggest standouts is that the platform generates both AI video and images, a step above some of the most competitive offerings on the market including OpenAI’s DALL-E 3, which only offers text-to-image generating capabilities in ChatGPT. 

Also: The best AI image generators: Tested and reviewed

OpenAI announced Sora, its text-to-video generator, in February but has yet to release the model to the public. Since then, several companies have released their generators, including Stability AI, Runway, and China-based Kling AI, while others, including Google, have announced that they are working on their video generators. 

Video generation is the next generative AI solution that companies are racing to perfect. Its practical applications include filmmaking, video game design, and content creation.