TLDR: Chinese AI firm ShengShu Technology has unveiled Vidu Q1, an advanced generative AI video model designed to compete directly with industry giants like OpenAI’s Sora and Google’s Veo. Launched in April 2025, Vidu Q1 boasts cinematic-grade visual effects, enhanced control, and multimodal capabilities, positioning it as a frontrunner in the rapidly evolving AI content creation landscape.
Beijing, October 21, 2025 – ShengShu Technology, a prominent Chinese artificial intelligence firm, has significantly escalated the global competition in generative AI video with the recent launch of its advanced model, Vidu Q1. Positioned as a direct rival to OpenAI’s groundbreaking Sora and Google’s Veo, Vidu Q1 represents a major leap in AI-driven content creation, offering sophisticated capabilities aimed at professional and consumer markets alike.
Unveiled in April 2025, Vidu Q1 is a multimodal AI model capable of generating highly realistic videos from both textual prompts and images. A key highlight of this new iteration is its enhanced ability to splice multiple entities—people, objects, and environments—into a single, coherent generated video, maintaining impressive consistency throughout. The model emphasizes improvements in quality and control, delivering cinematic-grade visual effects, natural transitions, and high-fidelity audio generation, all achievable with minimal input.
ShengShu Technology has not shied away from directly comparing Vidu Q1 to its international counterparts. The company highlighted that Vidu Q1 has achieved top rankings in VBench, an independent generative video evaluation standard, surpassing both OpenAI’s Sora and Google Gemini (which encompasses Veo) in performance metrics. This benchmarking underscores ShengShu’s ambition to lead the global generative AI video market.
The Beijing-based startup, founded in March 2023 by a team from Tsinghua University’s Institute for AI Industry Research, has rapidly evolved its Vidu platform. Earlier iterations, such as Vidu 1.0 launched in July 2024, quickly garnered over 10 million users and generated more than 100 million video clips, demonstrating the platform’s scalability and user adoption. Vidu 2.0, released in January 2025, further refined the generation process, creating 4-second, 512p videos in just 10 seconds and establishing logical relationships between multiple user-specified objects within a scene.
Vidu Q1 builds upon these foundations, leveraging a Universal Vision Transformer architecture that combines Transformer and Diffusion models—a design philosophy ShengShu claims to have explored even before OpenAI’s similar “Diversity in Transformation” architecture, with related research published in September 2022. The model is designed to generate scenes consistent with the laws of physics, featuring rich details such as realistic shadow effects and nuanced facial expressions.
ShengShu Technology’s commitment to innovation is further evidenced by its strategic partnerships. In March 2025, Vidu collaborated with Aura Productions, a Los Angeles-based animation studio, to develop a pioneering sci-fi anime series created entirely with generative AI, showcasing the model’s potential in professional content production.
Also Read:
- OpenAI’s Sora 2 Widens Reach Amid Mounting Concerns Over Child Safety and AI Misuse
- AI Revolutionizes Digital Art: Midjourney and Picflow Lead the Charge in Image and Video Generation
With Vidu Q1 now available to the public via its official website, ShengShu Technology is poised to significantly impact the generative AI video landscape, offering creators and enterprises powerful tools to produce high-quality, dynamic visual content and intensifying the race for AI supremacy.
