以下文章来源于智东西 ，作者江 宇 它让人们意识到，一个好的AI视频更需要创意调度。而过去，创作者在AI视频中最难掌握的两个要素，恰恰就是理解故事结构与镜头语言。 与此同时，另一款国产视频生成模型 生数科技的Vidu Q3 ，也在创作者社群中走红，并刚刚登顶全球权威AI评测平台Artificial Analysis，成为 全球排名 第一 的视频生成模型。 | Artificial Analysis | | Models ~ Speech, Image, Video ~ Hardware Leaderboards ~ Al Trends | | | Arenas · About · | | | For Enterprise | Insights Q Search ... WK | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | | Current models All models | No Audio With Audio | AH | Open weights | Global Leaderboard | | Personal ...

生数科技成立于2023年3月，是国内领先的多模态通用大模型研发企业。公司深耕视频生成及多模态内 容创作领域，构建了Vidu MaaS、Vidu SaaS、Vidu Agent的完整产品矩阵。其自主研发的Vidu视频生成 模型于2024年7月全球上线，首创"参考生视频"功能，率先攻克了商业视频制作中多主体连续一致性的 技术瓶颈。据Artificial Analysis榜单数据，Vidu生成速度较OpenAI Sora2快10倍，较Google Veo 3 Fast和 Grok-imagine-video快2倍，位居全球商业内容生成模型速度榜首。2025年12月，公司开源TurboDiffusion 框架，在单张RTX 5090显卡上仅需1.9秒即可生成5秒视频，将视频生成效率提升100-200倍。目前Vidu 已服务全球内容创作者及广告、动画、影视、教育、游戏、文旅等行业客户，2025年实现用户和收入超 10倍增长。 《商道创投网》创业家会员·本轮融资用途是什么？ 《商道创投网》2026年2月9日从官方获悉：生数科技近日完成了由中关村科学城公司、星连资本联合领 投，万兴科技、视觉中国、拓尔思战略投资，启明创投、 ...

商业化落地方面，生数科技构建了包含Vidu MaaS、Vidu SaaS、Vidu Agent、Vidu APP的完整应用矩 阵，用户覆盖全球200多个国家及地区，2025年实现用户与收入超10倍增长，其技术突破与商业化成果 正推动内容产业生产流程、生态格局重构，同时向物理世界的AI决策应用拓展。此次融资将进一步巩 固其技术壁垒与市场优势，加速AI生产力在全球产业中的规模化释放。 作为全球多模态生成算法领域先行者，生数科技于2022年提出的U-ViT架构早于OpenAI的DiT架构，核 心产品Vidu文生视频大模型全面对标Sora，首创"参考生视频"技术破解多主体连续一致性难题，近期发 布的Vidu Q3在Artificial Analysis榜单中位列中国第一、全球第二，生成速度较Sora2快10倍。2025年12 月，生数科技开源 TurboDiffusion 框架，在单张RTX 5090显卡上仅需1.9秒即可生成5秒视频，将视频生 成效率提升100-200倍。 2月5日，北京多模态生成技术创新企业生数科技宣布完成超6亿元人民币A+轮融资，此轮融资由中关村 科学城公司、星连资本领投，万兴科技、视觉中国等上 ...

Group 1 - The core viewpoint of the article is that Shengshu Technology has successfully completed over 600 million yuan in A+ round financing, with significant investments from various strategic partners and existing shareholders [1][2] - The financing round was led by Zhongguancun Science City Company and Xinglian Capital, with strategic investments from listed companies such as Wanxing Technology, Vision China, and Tuolisi [1] - Shengshu Technology's multimodal large model, Vidu, is set to launch globally in July 2024, addressing the challenge of multi-entity consistency in commercial video needs [1] Group 2 - Vidu has demonstrated a generation speed that is 10 times faster than OpenAI's Sora2 and twice as fast as Google Veo 3 Fast and Grok-imagine-video [1] - The company plans to open-source the TurboDiffusion framework by December 2025, achieving video generation efficiency improvements of 100-200 times on a single RTX 5090 graphics card [1] - Shengshu Technology has established an application matrix consisting of Vidu MaaS, Vidu SaaS, and Vidu Agent, empowering content creators and enterprises across various industries globally [2] Group 3 - Vidu has become a preferred model for commercial content creation among global creators, content organizations, and enterprises, with expectations of over 10 times growth in users and revenue by 2025 [2]

Core Insights - The article discusses the recent A+ round financing of Shengshu Technology, which raised over 600 million RMB, led by Zhongguancun Science City and Xinglian Capital, with strategic investments from several listed companies [2] - Shengshu Technology's multimodal model, Vidu, ranked first in China and second globally in the latest Artificial Analysis benchmark, surpassing competitors like Runway Gen-4.5 and Google Veo3.1 [2][3] - Vidu is set to launch globally in July 2024, introducing a unique "reference生视频" feature that addresses the challenge of multi-entity consistency in commercial video production [3][4] Financing and Investment - Shengshu Technology completed an A+ round financing of over 600 million RMB, with participation from existing investors including Qiming Venture Partners and Beijing AI Industry Investment Fund [2] - The financing round indicates strong confidence from both new and existing investors in Shengshu Technology's growth potential [2] Product and Technology - Vidu has established an application matrix including Vidu MaaS, Vidu SaaS, and Vidu Agent, catering to various industries such as advertising, animation, education, and gaming [4] - Vidu boasts the fastest content generation speed globally, generating videos 10 times faster than OpenAI's Sora2 and twice as fast as Google Veo 3 Fast [3] - The company plans to open-source the TurboDiffusion framework in December 2025, significantly enhancing video generation efficiency [3] Market Position and Growth - Vidu has become a preferred model for commercial content creation among global creators and enterprises, with expectations of over 10 times growth in users and revenue by 2025 [4]