Shengshu Technology, a generative AI company born out of Tsinghua University’s AI research labs, has officially launched Vidu Q4 Preview, the latest version of its Vidu AI video generation family.
Shengshu Technology has launched Vidu Q4 Preview, the latest version of its Vidu AI video generation family.
The model is designed for short drama production companies, advertising agencies, social content creators and film production teams.
New features include reference-voice support, cinematic 4K video with 10-bit color and improved realism and lighting.
Vidu Q4 Preview supports up to 15 reference images to build complex scenes and concepts.
Pricing starts at $0.045 per second of video at 540P, up to $0.39 per second at 4K.
Image-to-video and reference-to-video generation is available at a 30% discount through November 30.
Available now, Vidu Q4 Preview is designed to deliver truly lifelike expressiveness. Combining cinematic image quality with faster generation, Vidu Q4 Preview is built for professionals who produce video at scale, including short drama production companies, advertising agencies, social content creators and film production teams.
“What impressed me most was the performances. Even in a pause with no dialogue, the characters' eyes and expressions alone were enough to convey the emotion beautifully. The visual effects are remarkably consistent,” Youjieunknown, a video creator who has worked with Vidu Q4 Preview. "The light from explosions illuminates the surrounding buildings, and the smoke and dust spread along with the collapsing structures. It looks like a real live-action shoot. I also tested how it handles text within the frame. Q4 is really strong at rendering Chinese text — on par with what you'd get from doing your own post-production visual effects work.”
More authentic performances with reference-voice support keeping a character’s voice consistent, with natural emotional delivery and more compelling acting
Cinematic 4K video with 10-bit color, so footage holds up in professional workflows and across formats
Improved realism and lighting with natural texture and depth
Stable camera moves and cuts that follow the creator's direction
Richer creative control allows for up to 15 reference images to be combined to build complex scenes and concepts
Cost-effective creation, with faster turnaround and high quality output at a lower production cost
Vidu Q4 Preview is available now: https://platform.vidu.com/. Pricing starts at $0.045 per second of video at 540P and scales with resolution, up to $0.39 per second at 4K. Image-to-video and reference-to-video generation with Vidu Q4 Preview is available at a 30% discount through November 30.
About Shengshu Technology
Shengshu Technology is a leading Chinese generative AI company and the creator of Vidu, a cutting-edge AI video generation model. Founded by a team from Asia's top-ranked Tsinghua University, it is building the world's first general world model to unify the digital and physical worlds. Guided by first-principles thinking, it uses visual and auditory information to train its foundation world model, enabling modeling, reasoning, predicting, and acting upon the rules of both worlds, and ultimately AGI. Its Vidu video model has grown into a global platform, and public reports show the company has raised nearly RMB 6 billion (approximately $890 million USD) since February 2026.
