突破三大技术瓶颈，极佳视界推出高可用世界模型体系。 当AI能力开始与物理世界连通，"具身智能"成为落地载体。 今年以来，"世界模型"在具身领域快速升温：谷歌、OpenAI、特斯拉、英伟达等巨头密集布局。 多位业内人士判断，世界模型将会缓解具身智能在数据稀缺与泛化困难上的瓶颈，极可能在VLA之 后，成为2026年的核心技术趋势。 在这一背景下，专注世界模型方向的物理AI公司——极佳视界，在近两个月内连续完成三轮融资，并 宣布最新进展。 据《智能涌现》获悉，极佳视界近日完成新一轮亿元级A1轮融资，本轮融资由华为哈勃、华控基金联 合投资。此前8月底，极佳视界宣布完成Pre-A&Pre-A+连续两轮数亿元融资。 通俗讲，世界模型是在数字世界里建模物理世界和运行规律：让AI在动手前，先在"脑内"搭建简化的物 理沙盒，预测下一秒会发生什么并据此选择执行动作，从而在陌生环境中少试错、更稳健。 2个月3轮融资，体现了资本市场对极佳视界团队实力、技术路线和业务推进的认可，也折射出投资方 对"物理世界通用智能"（物理AI）关键转折点的判断。 成立于2023年的极佳视界，聚焦物理AI，专注于"世界模型驱动的物理世界通用智能"。其产 ...

Core Viewpoint - The article highlights the successful completion of several rounds of financing by Beijing Jiajia Vision Technology Co., Ltd. (referred to as "Jiajia Vision"), a leading domestic company in the field of Physical AI, amounting to several hundred million yuan in Pre-A and Pre-A+ rounds, indicating a growing interest and investment in the Physical AI sector [2][4]. Financing Overview - Jiajia Vision completed two rounds of financing on August 28, 2025, raising several hundred million yuan, with the Pre-A round led by Guozhong Capital and followed by Zifeng Capital and PKSHA Algorithm Fund, while the Pre-A+ round was backed by CICC Capital, Guangzhou Industrial Investment, Yicun Songling, and Huqiang Capital [2][3]. - The company has now completed a total of six financing rounds, with the last round prior to this being a several tens of millions yuan angel round in February 2025 [4]. Company Background - Founded in June 2023 and based on the intelligent vision laboratory of Tsinghua University, Jiajia Vision initially focused on spatial intelligence but has since shifted to Physical AI, specializing in "world model platforms and embodied foundational models" [5][6]. - The company aims to accelerate the development of general intelligence in the physical world through its products, which include the GigaWorld world model platform and GigaBrain embodied foundational model [6][11]. Technology and Product Development - Jiajia Vision's products are designed to enable robots, autonomous vehicles, and intelligent spaces to perceive, understand, and execute complex operations in the real world, marking a significant advancement in the Physical AI field [6][12]. - The GigaBrain-0 model, released in July 2025, utilizes over 90% of its training data generated from Jiajia Vision's self-developed world model platform, showcasing a significant efficiency advantage over traditional data collection methods [12]. Market Position and Collaborations - The company has established partnerships with leading enterprises in various sectors, including intelligent driving and embodied intelligence, to facilitate large-scale industrial applications [9][18]. - Jiajia Vision is recognized as the first domestic startup focusing on world models and is positioned at the forefront of this emerging field [6][17]. Leadership and Team - The core team includes experienced professionals with backgrounds in AI and robotics, such as the founder and CEO Huang Guan, who has over ten years of experience in AI technology and industry [10][11].

Core Insights - Physical AI company, GigaVision, has successfully completed two rounds of financing totaling several hundred million yuan, indicating strong market recognition of its team, technology, and product progress [2] - The company aims to accelerate towards general intelligence in the physical world through its world model-driven foundational models [2][3] Financing and Investment - GigaVision completed Pre-A and Pre-A+ financing rounds led by various investment firms, including Guozhong Capital and CICC, showcasing investor confidence in its capabilities [2] - The company has raised significant funds in multiple rounds, including a recent angel round, reflecting robust investor interest [2] Technology and Product Development - GigaVision focuses on "world model-driven physical world foundational models," with products like GigaWorld and GigaBrain, which are designed to enhance physical AI capabilities [2][10] - The company is positioned as a leader in the world model and VLA (Vision-Language-Action) model sectors, with ongoing collaborations with major automotive and robotics companies [5][16] Market Position and Vision - GigaVision is recognized as the first domestic company specializing in world models, aiming to lead technological advancements and industry applications [5] - The company envisions a "ChatGPT moment" for the physical world within 2-3 years, predicting significant breakthroughs in physical AI technology and applications [3][4] Team and Expertise - The core team comprises top researchers and industry experts from prestigious institutions, contributing to GigaVision's strong research and development capabilities [6][7] - The leadership team has extensive experience in AI and has achieved numerous accolades in global AI competitions, enhancing the company's credibility [6][7] Strategic Partnerships and Collaborations - GigaVision has established partnerships with leading automotive manufacturers and robotics centers to facilitate large-scale production and application of its technologies [16] - The company is actively pursuing collaborations to explore the deployment of embodied intelligence across various sectors, including industrial and consumer markets [16][20] Future Outlook - GigaVision plans to utilize the recent funding for technology development and market expansion, aiming to enhance customer delivery and accelerate towards the physical world ChatGPT moment [17] - The company is committed to achieving world-class technological breakthroughs in world models and embodied intelligence, with a focus on creating social value [21]