Baidu Intelligent Cloud and Shengshu Technology have joined forces to officially launch Vidu, China’s first fully self-developed large-scale video model! This AI artifact that benchmarks against OpenAI Sora, with its powerful Wensheng video and Tusheng video functions, up to 1080P resolution, and amazing inference speed (it only takes 30 seconds to generate a 4-second video), is ideal for art design , game production, film and television animation and other creative fields have brought revolutionary changes. Vidu not only has core advantages such as high dynamics, high fidelity, and high consistency, but also innovatively adds special capabilities such as character consistency, animation style, text and special effects screen generation, etc., aiming to comprehensively improve human productivity and creativity, and open up A creative future.
Vidu, a large video model jointly launched by Baidu Smart Cloud and Shengshu Technology, is officially launched, bringing new possibilities to the creative field. Vidu is the first fully self-developed large-scale video model in China. Its performance is fully aligned with OpenAI's Sora. It is oriented to creative fields such as art design, game production, film and television animation, and social entertainment, and is committed to stimulating the productivity and creativity of all mankind.
Vidu has two functions: Wensheng video and Tusheng video, providing two video duration options of 4 seconds and 8 seconds, with a resolution up to 1080P. On the basis of maintaining high dynamics, high fidelity and high consistency, it adds new features such as character consistency, animation style, text and special effects screen generation. Vidu's inference speed is so fast that the user's waiting time to generate a 4-second video clip is only 30 seconds, which is twice as fast as the fastest in the industry, bringing users a near-seamless creative experience.
Baidu Smart Cloud provides Vidu with systematic, end-to-end large model training services, including large-scale computing power support and full-link guarantee. Vidu adopts Baidu Baige AI heterogeneous computing platform to achieve "fast, stable and economical" large model development, and the effective training time reaches 98.8%.
With the official launch of Vidu, Shengshu Technology and Baidu Smart Cloud look forward to promoting AI to empower the creative future through technological innovation, uncovering multi-modal models for more scene applications, and leading us into a more efficient, convenient and creative world. digital world.
The launch of Vidu marks a new milestone in domestic AI video generation technology. Its efficient and convenient creative experience will greatly unleash creative potential and bring unlimited possibilities to all walks of life. In the future, we expect Vidu to play a role in more fields, lead AI technology innovation, and create a better digital world.