The global AI video generation race has a new major contender as Chinese startup Shengshu AI unveiled Vidu Q2, a sophisticated new model designed to directly challenge OpenAI’s Sora. The launch signals intensifying competition to create realistic and controllable AI-generated video content for creators and enterprises worldwide.
Vidu Q2’s Unique ‘Reference-to-Video’ Feature
Vidu Q2’s standout capability is its “Reference-to-Video” feature, which allows creators to produce highly consistent and expressive videos using up to seven reference images. These images can define specific faces, gestures, scenes, or props, which the model then merges into a single, coherent clip. According to Shengshu AI, this improves realism, cinematic motion, and prompt comprehension, moving beyond basic video creation to help AI “act and tell stories alongside creators.” The model supports text-to-video and image-to-video generation, producing clips of up to eight seconds with synchronised dialogue and sound effects.
Global Ambitions and Competitive Landscape
Shengshu AI is making Vidu Q2 and its API globally available through its platform, offering membership tiers ranging from a free plan to a flagship option costing approximately $84 per month. This move places the Beijing-based company in direct competition with major international players like OpenAI’s Sora and Google DeepMind’s Genie and Veo models. The field is also crowded within China, with tech giants like Kuaishou, ByteDance, Alibaba, and Tencent developing their own powerful video generation tools, highlighting the strategic importance of this technology.
What This Means For The MENA Tech Scene
The launch of another powerful, globally accessible AI video model has significant implications for the MENA region. For the burgeoning creator economy and digital marketing industries in hubs like Dubai and Riyadh, tools like Vidu Q2 offer new, cost-effective ways to produce high-quality video content. Furthermore, the rapid innovation from global competitors serves as both a benchmark and a catalyst for MENA’s own AI startups. It underscores the market opportunity for developing AI models that are finely tuned to regional languages, cultural nuances, and specific industry needs, potentially attracting further investment from local VCs keen on backing the next wave of AI innovation.
About Shengshu AI
Founded in March 2023, Shengshu AI is a Beijing-based company focused on developing multimodal AI models. The company’s leadership team has strong ties to the prestigious Tsinghua University. Despite being a young company, Shengshu AI has already attracted significant investment, recently closing a Series A funding round of hundreds of millions of yuan. Its backers include prominent names such as Baidu, Qiming Venture Partners, Alibaba-affiliated Ant Group, and the Beijing municipal government-backed AI Industry Fund.
Source: South China Morning Post
