MiniMax, the latest AI video generator from China, is turning heads with its ability to create hyper-realistic human footage, including intricate hand movements—a challenge that has stumped many other AI tools.

This innovation is the latest breakthrough from the unicorn startup, MiniMax, which is backed by tech giants Alibaba and Tencent. MiniMax is already known for its AI companion app, Talkie, which has seen over 15 million downloads, offering users the ability to converse with virtual characters similar to Character.ai.

The official demo shared on X (formerly Twitter) showcases a stunning example: a child embarks on a magical journey through history by touching a coin, with realistic characters and special effects—all generated from simple text prompts using AI and advanced editing techniques. To test the tool, I created an account and experimented with various prompts. While MiniMax is indeed impressive and competitive with other AI video generators like Runway Gen-3, Dream Machine, and Kling, it doesn't quite surpass the hype created by its demo video.

MiniMax is the latest addition to the startup's suite of generative models, which also includes AI tools for speech, language, and music. Although the video model was launched quietly, it quickly gained attention across social media in both China and the West. Yan Junjie, the founder of MiniMax, stated that their internal evaluations show superior performance compared to Runway in video generation. The company is already working on version-02, which promises to include features like image-to-video and extended clip generation, pushing the boundaries of what AI can achieve in video creation.