Kuaishou, a major player in China's short video market, is known for its innovative approaches to video content creation. With over 400 million daily active users, it is second only to ByteDance's Douyin. Kuaishou has recently launched a new AI model, Kling, designed to generate high-quality videos from simple text prompts, putting it in direct competition with other AI video generation tools like OpenAI's Sora.

Key Update

Kuaishou's latest AI model, Kling, can transform text prompts into stunning 1080p video clips up to two minutes long. Currently in its trial phase and accessible through a waitlist, Kling supports various aspect ratios and can generate both realistic and imaginative scenes, such as a white cat driving through city streets or a boy eating a cheeseburger. Using a diffusion transformer model, similar to OpenAI's Sora, Kling produces videos with realistic physics and smooth movements. It also boasts advanced 3D face and body reconstruction, enhancing expression accuracy and limb movements. This addition to Kuaishou's AI innovations, alongside other tools like the KwaiYii large language model and the Kolors text-to-image model, showcases the company's continuous push for cutting-edge AI developments.

Technical Terms

AI Model

A program or system trained to perform tasks by learning patterns from data.

Text Prompt

A textual input given to an AI model to generate content.

Diffusion Transformer Model

A type of machine learning model used for generating high-quality images or videos by simulating how particles spread out over time.