Jijia Shijie completed a Pre-Series B financing of 1 billion yuan, driven by the "world model" to build the world's leading "embodied schema".

Recently, "Giga Vision", a leading enterprise in embodied schemas and general-purpose robots, announced the completion of a nearly 1 billion RMB Pre-B round of financing. The investors in this round include top chip and automotive industry capitals such as SMIC Juyuan, Shanghai Semiconductor Industry Investment Fund, Linxin Capital, Xingyuan Capital, and Wanlin International, as well as significant state-owned capital platforms and well-known financial institutions like CICC Capital, Suzhou Venture Capital, Huaqiang Capital, Yangtze River Capital, Optics Valley Industrial Investment, Xishan State-owned Investment, Jinyu Maowu, Xinding Capital, Lingyang Investment, Caixin Capital, Zhangke Yaokun, and Chengzhu Investment. Among them, old shareholders such as CICC Capital, Huaqiang Capital, Caixin Capital, and Zhangke Yaokun continued to provide substantial oversubscribed support. Gengxin Capital China served as the financial advisor.

Giga Vision is a leading domestic company in "embodied schemas and general-purpose robots". With the four-in-one concept of "embodied schema - world model - native ontology - generalized scenarios", it promotes the entry of general-purpose robots into thousands of households and industries, leading the explosion of AGI in the physical world.

Core Advantages

1. Leading "Embodied Schema" in the Industry

The GigaBrain series of embodied schemas are at the global leading level. The open-source model GigaBrain-0.1 once won the first place in the authoritative large-scale real-machine evaluation globally. The GigaBrain series are all at the leading level in the industry in terms of the comprehensive capabilities of embodied schemas, such as ten-thousand-hour open-source schemas, basic model technology reports, real-machine complex task effects, and large-scale authoritative real-machine evaluation results.

2. The First in China to Layout the "World Model", with Dual Values of Model Architecture and Data Source, Aiming to Bring a 10 - 100 Times Efficiency Improvement

In terms of model architecture, GigaBrain-0.5M* is the world's first embodied schema based on the world model to achieve efficient learning and self-evolution through reinforcement learning. GigaWorld-Policy is the first in the world to achieve a comprehensive breakthrough in real-time performance, training efficiency, and success rate of the "world - action model WA".

In terms of data source, the GigaWorld and DriveDreamer series continue to lead the development of embodied and driving world models as a new generation of simulators.

3. Full-stack Layout of "Native Ontology" and "Generalized Scenarios"

In terms of native ontology, the full-stack self-developed native ontology Maker H01 of Giga Vision has started large-scale mass production and delivery. More models of native ontologies are about to be released, targeting multiple scenarios such as data collection, industry, commerce, and households. It aims to deliver a thousand units throughout the year.

In terms of generalized scenarios, Giga Vision has achieved pioneering application implementations for multiple benchmark customers in multiple scenarios such as automobile manufacturing, 3C electronics, warehousing and logistics, advanced guided tours, and household terminals.

"The OpenAI of the Physical World"

Industrial giants and top capitals are collectively betting on it.

Digital AGI represented by language models is reshaping the entire digital world. 50% of the global GDP still lies in the physical world. In the future, most of it will be dominated by the new industrial chain brought by physical AGI. The key to the explosion of physical AGI lies in the breakthrough of the basic model in the physical world. With the rapid development of general models in the physical world such as end-to-end, VLA, and world models, physical AGI is on the verge of an explosion.

In terms of capital layout, Giga Vision has gathered a powerful capital lineup of "top industrial capitals + leading financial institutions + significant state-owned capital platforms", which is a concentrated manifestation of the entry of physical AGI into the explosive stage. Global top industrial capitals such as Huawei Hubble, SMIC Juyuan, BAIC Industrial Investment, and PKSHA have provided important support for the company's strategic layout and industrial collaboration. The recognition of many leading financial institutions such as CICC Capital, Fortune Capital, Huakong Fund, and Miracle Plus fully reflects the industry status of Giga Vision. The support of many significant state-owned capital platforms such as Suzhou Venture Capital, Guangzhou Industrial Investment, Yangtze River Capital, and Zhuhai Science and Technology Industry Group has provided long-term capital and ecological support.

Giga Vision is the first domestic technology company to layout the world model. Since its establishment, it has focused on "the basic model of the physical world driven by the world model". With the collective bets of industrial giants and top capitals, Giga Vision has rapidly grown into a leading domestic company in "embodied schemas and general-purpose robots", aiming to become "the OpenAI of the physical world". With the four-in-one concept of "embodied schema - world model - native ontology - generalized scenarios", it promotes the entry of general-purpose robots into thousands of households and industries, leading the explosion of AGI in the physical world.

"The World Model Brings a 10 - 100 Times Efficiency Improvement"

The core team of Giga Vision was the first in China to conduct research and development on end-to-end models in the physical world, and has continuously led the technological breakthroughs and industrial implementations of physical AI algorithms such as object recognition, face recognition, BEV, OCC, end-to-end, and world models.

In the field of embodied intelligence, Giga Vision released the GigaBrain series of embodied basic models, published the GigaBrain-0 technical report, and open-sourced basic models such as GigaBrain-0 and GigaBrain-0.1, achieving real-machine effects for long-range complex tasks. In the currently largest real-machine evaluation competition globally, the open-source model GigaBrain-0.1 surpassed many models such as Pi0.5 and won the first place globally.

While continuously making breakthroughs in embodied schemas, Huang Guan, the founder and CEO of Giga Vision, said:

"We must recognize that there are two key problems with the current VLA-dominated embodied schemas: one is the too-low efficiency of the model architecture, and the other is the too-low efficiency of real data collection.

The rapid development of the world model provides an unprecedented, pioneering, and fundamental solution to the problems of 'model architecture' and 'data source'."

Giga Vision is the first company in China to layout the "world model". It has in-depth accumulations in both the application of the world model to "model architecture" and "data engine", aiming to bring a 10 - 100 times efficiency improvement through the world model.

In terms of model architecture, Giga Vision released GigaBrain-0.5M*, which is the world's first embodied schema based on the world model to achieve efficient learning and self-evolution through reinforcement learning. Giga Vision released GigaWorld-Policy, which is the first in the world to achieve a comprehensive breakthrough in real-time performance, success rate, and training efficiency of the "world - action model WA", enabling the "world - action model" to truly enter the large-scale scaling stage.

In terms of data engine, Giga Vision continues to lead the technological breakthroughs and industrial implementations of embodied and driving world models as a new generation of simulators.

In the embodied direction, Giga Vision released GigaWorld-0, which is the first in the world to let the embodied world model play a core role in high-level embodied schemas. GigaWolrd-0.5, as the basic model of GigaWorld-Policy, has brought unprecedented improvements in efficiency and effectiveness.

In the driving direction, the series of work such as DriveDreamer and DriveDreamer4D of Giga Vision are the earliest series of pioneering work globally to apply the world model in the physical world, and have started large-scale industrial implementation and application.

"Native Ontology" Starts Large-scale Mass Production and Delivery

Pioneering Application Implementations for Multiple Benchmark Customers in "Generalized Scenarios"

The development of embodied schemas also requires the collaboration of native ontologies. In November 2025, Giga Vision released the full-stack self-developed new-generation physical AGI native ontology Maker H01, making many pioneering design optimizations in terms of human-machine collaboration, cost-effectiveness, function priority, and data priority. Maker H01 has currently started large-scale mass production and delivery, targeting multiple scenarios such as data collection, industry, and services. Giga Vision will release more models of native ontologies suitable for different scenarios in 2026, aiming to deliver a thousand units of ontologies throughout the year.

In terms of scenarios, Giga Vision focuses on building general-purpose robots for high-value and highly generalized scenarios based on embodied schemas. Currently, it has reached close cooperation with leading customers in multiple scenarios such as automobile manufacturing, 3C electronics, warehousing and logistics, guided tours, and household services, accelerating the entry of general-purpose robots into thousands of industries and households.

The Dream Team of Basic Models for Physical AGI

Pioneers in Industrial Implementation

Giga Vision closely relies on the Intelligent Vision Laboratory of the Department of Automation at Tsinghua University. The core team has witnessed the development of physical AI in the past decade, including CV, autonomous driving, embodied schemas, and world models, and has achieved world-class results in technological innovation and industrial implementation at each stage.

Huang Guan, the founder and CEO of Giga Vision, is a Doctor of Innovative Leading Engineering from the Department of Automation at Tsinghua University. He once served as the head of vision perception technology at Horizon Robotics, a partner and vice president of algorithms at Jianzhi Robotics, and has work experience at top research institutions such as Microsoft Research Asia and Samsung China Research Institute.

Dr. Huang Guan has witnessed the technological and industrial development of physical AI in the past decade. He has led the team to win world championships in many global authoritative AI competitions and published many globally well-known AI achievements. He has rich experience in technological innovation, industrial implementation, and continuous entrepreneurship in the field of physical AI.

The core model team of Giga Vision is one of the few in the industry that has both the top "comprehensive technological frontier innovation ability of the new generation of physical AI" and the "full-stack system mass production experience of traditional physical AI". The core members include former vice presidents and architecture experts of large Internet companies, top 2% scientists in the world with nearly 20,000 citations, top scientists in the world model with more than 10 first-author papers in top conferences during their doctoral studies, industrial experts with more than 10 years of full-stack mass production experience in physical AI, Huawei's "Genius Youth" awardees, and top algorithm and infrastructure experts in linear acceleration of ten-thousand GPU clusters. They can be regarded as the dream team of basic models for physical AGI.

In the research and development direction of humanoid ontologies, Giga Vision has gathered the R & D leaders of domestic top robot enterprises with high shipment volumes, and has cumulatively led the R & D and delivery of thousands of humanoid robot products. The team is well-versed in the whole process of robot ontology design, performance optimization, and mass production, laying a solid core foundation for the physical implementation of embodied schemas.

Giga Vision also has an industrial team with rich implementation experience, including industrial executive expert teams from technology giants such as Horizon Robotics, Alibaba Cloud, Bosch, and Baidu. They have more than 20 years of cumulative experience in leading implementations worth more than 3 billion in the field of physical AI. Giga Vision aims to move towards the new generation of super terminals for physical AGI based on embodied schemas and native ontologies.

This round of financing will be used to continuously lead the evolution of embodied schemas and world models, and at the same time promote the adaptation of native ontologies and the implementation of generalized scenarios, aiming to reach the GPT-3 moment of embodied schemas with the highest efficiency, fastest speed, and strongest performance, leading the explosion of AGI in the physical world.

This article is from the WeChat official account "GigaAI Technology", author: GigaAI Technology. It is published by 36Kr with authorization.