Morning Star | Multiple Qiming Venture Partners Portfolio Companies Appear at 2025 Zhongguancun Forum Annual Conference and Win Multiple Awards

Since its founding, Qiming Venture Partners has focused on investing in exceptional early- and growth-stage companies driven by technology and innovation.

From March 27 to 31, the 2025 Zhongguancun Forum Annual Conference was held in Beijing. This year's conference took "New Quality Productive Forces and Global Sci-Tech Cooperation" as its theme, featuring 128 events across five major sections: forum sessions, technology trading,成果发布, frontier competitions, and supporting activities. Centered on frontier fields such as AI large models, embodied intelligence, quantum technology, biomedicine, and brain-computer interfaces, participants engaged in in-depth discussions on cutting-edge technological advances and industry development trends.

Since its founding, Qiming Venture Partners has focused on investing in outstanding early- and growth-stage companies driven by technological innovation. At this year's forum, portfolio companies including Galaxy Universal, Zhipu AI, Shengshu Technology, Infinigence AI, Mech-Mind Robotics, Yunji Technology, and DeepWise made appearances, demonstrating their technological capabilities through keynote speeches, partnership announcements, model releases, and product showcases. Xiaomi (01810.HK), Zhipu AI, Galaxy Universal, Shengshu Technology, Infinigence AI, DeepWise, StepFun, and Thorough Future received multiple recognitions.

01/

He Wang of Galaxy Universal Deeply Engaged at Zhongguancun Forum

Galbot Shows a New Vision for Future Robot Work

He Wang, founder and CTO of Galaxy Universal, was deeply involved in this year's forum. He was invited to deliver a keynote address titled "Ushering in the Productivity Era of Humanoid Robots" at the opening plenary session, providing theoretical support for the synergistic development of technological breakthroughs and industrial deployment. Dr. Wang noted that as technological waves continue to surge forward, the next era for humanoid robots will be the "productivity era," in which robots will work autonomously and generalize operations, driven by end-to-end embodied large models based on closed-loop visual input, enabling cross-industry applications of general-purpose robots and delivering new quality productive forces to society.

Dr. Wang was also invited to participate in the AI Youth Innovation and New Quality Productive Forces Development Conference, the Future AI Pioneer Forum, and the Beijing International Youth Innovation Development Forum, focusing on embodied intelligence and engaging with domestic and international scholars and enterprise representatives to explore the future direction of this field.

At the AI-themed special forum — the AI Leading Talent Development Forum — Galaxy Universal joined forces with industry, academia, and research institutions to establish an Embodied Intelligence Joint Laboratory, with a signing and launch ceremony held at Beijing Zhongguancun Academy. The joint laboratory will adopt an integrated "learning-research-production" cooperation model, focusing on frontier research and application innovation, cultivating high-end talent, facilitating technological innovation and industrial upgrading, and building an outstanding embodied large-model robot R&D base.

In the core exhibition area of the forum's main venue, the "Xiaoguan Store" operated by Galaxy Universal's embodied large-model robot Galbot became a focal point. After audience members placed orders, Galbot would accurately identify and pick up items, then autonomously plan its route to deliver them. This retail unmanned operation experience space painted a new picture of future robot work. With only simple deployment, Galbot stably and excellently completed thousands of tasks over five consecutive days, with a task success rate exceeding 99.97%.

02/

Zhipu AI Releases AutoGLM Rumination

Integrating Deep Research and Practical Operation Capabilities

At the Zhongguancun Forum, Zhipu AI officially released AutoGLM Rumination, a new intelligent agent that not only possesses deep research capabilities (Deep Research) but can also perform practical operations (Operator), truly pushing intelligent agents into a "thinking while doing" phase.

The model behind AutoGLM Rumination is Zhipu AI's newly introduced agent brain — the Rumination model, which uses reinforcement learning to enable the model to learn self-criticism, reflection, and even rumination, trading longer deep-thinking time for superior results. The Rumination model achieves real long-horizon reasoning and task execution by breaking through real-time web search, dynamic tool calling, deep analysis, and self-verification.

Notably, this is the world's first agent integrating deep research and practical operation capabilities, marking an important advance in autonomous agent technology and a further upgrade of device-control agents. The core chain models and technologies of AutoGLM Rumination will be officially open-sourced on April 14 to promote industry ecosystem development.

Based on Zhipu AI's self-developed full-stack large model technology, AutoGLM Rumination integrates the general capabilities of GLM-4, the reflection capabilities of GLM-Z1, the rumination capabilities of GLM-Z1-Rumination, and the automatic execution capabilities of AutoGLM. In terms of reasoning models, Zhipu AI introduced more reasoning-oriented data and deeply optimized general capabilities during the alignment phase, launching the new deep-thinking model GLM-Z1-Air, whose performance is comparable to DeepSeek-R1 (671B, 37B active).

Zhipu AI has always focused on AGI foundation model R&D and has currently advanced to the L3-Agentic LLM stage. Regarding industry ecosystem development, Zhipu AI insists on co-creation with industry partners, using its accumulated expertise in large model R&D to help partners jointly develop successful large model applications. Zhipu AI also actively promotes the global expansion of Chinese original large models and solutions.

03/

Shengshu Technology's Jun Zhu Releases Multimodal Video Generation Model

and Delivers Keynote Speech

Recently, Jun Zhu, Vice Dean of Tsinghua University's Institute for Artificial Intelligence, founder and chief scientist of Shengshu Technology, released the multimodal video generation model Vidu Q1. This model focuses on "high controllability," aiming to solve the "randomness" pain point in current AI video generation. The model will launch globally in April.

Vidu Q1 achieves significant breakthroughs in multi-subject detail controllability (particularly motion and layout control), audio-visual synchronization control, and image quality enhancement. Additionally, Vidu Q1 can further enhance image quality, delivering clearer, more refined, and more realistic visual experiences to users.

In his keynote speech titled "China's Global Leadership in Multimodal Generation," Professor Zhu emphasized the key role and repeated technological breakthroughs of domestic video large models in global technological competition. He also shared Shengshu Technology's forward-looking development path for Vidu based on practical exploration and predictions, clearly showing more possibilities for multimodal technology in virtual-real integration and interaction with the physical world. The release of Vidu Q1 not only demonstrates China's deep technical accumulation in video large models but also showcases the leading strength of domestic large models on the global stage.

04/

Lixue Xia of Infinigence AI Attends Global Unicorn Enterprise Conference

Computing Ecosystem Platform Featured in Innovation Exhibition

During the Zhongguancun Forum, Lixue Xia, co-founder and CEO of Infinigence AI, was invited to the 2025 Global Unicorn Enterprise Conference to discuss AI industry development trends with global innovation leaders. Xia noted that future Chinese unicorn enterprises will leverage solid technical accumulation and innovative industry systems to unleash unique and tremendous development potential, continuously achieving major breakthroughs in core areas of global technological development and contributing Chinese solutions to global industrial intelligent upgrading.

At the same time, Infinigence AI participated in the launch ceremony of the Beijing AI Public Computing Platform ecosystem network construction, which will provide more accessible and open public computing services to universities, research institutes, and AI enterprises in the Haidian area.

During the forum, Infinigence AI's computing ecosystem platform was also showcased as an AI innovation exhibit. This platform addresses the problems of "unavailable," "unusable," and "unaffordable" computing resources that SMEs face, and is generating positive radiating effects nationwide. It has already established benchmark demonstration cases in Shanghai's Xuhui district and Hangzhou, Zhejiang, providing replicable and scalable experience for regional computing resource market-oriented operations and intensive utilization through computing public service platform construction.

05/

Mech-Mind Robotics Showcases at Zhongguancun Forum Tech Exhibition

"Hand-Eye-Brain" Integration Demonstrates Powerful Generalization

At this year's forum tech exhibition, Mech-Mind Robotics displayed its latest Mech-GPT robot brain and Mech-Hand multi-finger dexterous hand, paired with Mech-Eye robot vision. The products demonstrated strong generalization capabilities, handling highly diverse tasks and recognizing and manipulating a wide variety of objects.

To date, Mech-Mind Robotics has deployed over 15,000 intelligent robot vision devices in more than 50 countries and regions worldwide. By equipping robots with intelligent "eyes" and "brains," Mech-Mind Robotics makes robots more capable intelligent assistants, bringing intelligent and flexible upgrades to automotive, logistics, construction machinery, new energy lithium battery, and other sectors.

06/

Yunji Technology Reaches Partnership with Hong Kong Norda Robotics

During the parallel forum — the Beijing-Hong Kong Sci-Tech Innovation Forum — at this year's conference, Yunji Technology and Hong Kong Norda Robotics signed a strategic cooperation agreement around joint R&D and market development.

In 2024, Yunji Technology established its Hong Kong branch and invested in a Hong Kong R&D center. Through deep industry-academia-research-application integration, it aims to build strong Chinese AI technology and strong Chinese robot brands, using Hong Kong as a window to lead and apply globally.

07/

DeepWise and AI for Science Institute, Beijing Jointly Release

Upgraded Bohrium Research Space Station

At this year's forum Youth Science Forum, DeepWise and the AI for Science Institute, Beijing (AISI) officially released a new version of the Bohrium research space station.

This AI for Science flagship platform, named after quantum mechanics pioneer Niels Bohr, has been called "the Hugging Face of the research world." As the world's first AI research platform covering "literature reading — computation — experimentation — multidisciplinary collaboration," Bohrium features the newly upgraded research knowledge base and AI academic search platform "Science Navigator (SN)" at its core, helping to restructure research processes and unleash research creativity.

With advanced artificial intelligence technology at its core, the platform systematically addresses key pain points researchers face in literature screening, cross-disciplinary knowledge discovery, and academic resource access, particularly effectively tackling challenges of information overload, cumbersome retrieval, and time-consuming organization. Facing the realities of massive, rapidly updating, and diversely formatted literature, Science Navigator provides global researchers with "one-stop" intelligent research support, significantly improving research efficiency and facilitating scientific breakthroughs.

08/

Multiple Qiming Venture Partners Portfolio Companies Selected

for 2025 Zhongguancun Forum Major Sci-Tech Achievements

At the opening ceremony of the Zhongguancun Forum, ten major sci-tech achievements were released. The "Beijing AI Innovative Application Achievements," jointly released by Xiaomi, Zhipu AI, Galaxy Universal, Shengshu Technology, and other institutions, as well as the "Beijing Major Open-Source Achievements" jointly released by Zhipu AI, were selected.

Xiaomi's achievement was "AI-Empowered High-Performance Alloy Material R&D at Xiaomi," which improved R&D efficiency by several dozen times, rapidly identifying new die-casting solutions with optimal performance and environmental benefits. Zhipu AI's achievement was "AI Tools Bringing Personalized Learning into Reality"; based on the powerful capabilities of its fully self-developed GLM foundation large model, Zhipu AI has built a complete suite of model products benchmarked against advanced international standards, including the AI productivity assistant Zhipu Qingyan, the AI video generation product Qingying, and the large model intelligent interaction agent AutoGLM. Galaxy Universal's achievement was the "Humanoid Robot Smart Pharmacy Solution," in which a smart pharmacy is operated 24/7 unmanned by the embodied large-model robot Galbot. Leveraging end-to-end embodied large models, end-to-end navigation, and other advanced key technologies, this solution addresses traditional pharmacy challenges of insufficient staffing, limited service hours, and high operating costs, while providing a replicable demonstration sample for the intelligent retail industry with broad promotion value in commercial retail, industrial, and logistics sectors. Shengshu Technology's Vidu is China's first fully self-developed, long-duration, high-consistency, high-dynamics video large model, currently available globally. Vidu supports three core functions: text-to-video, image-to-video, and reference-to-video, with leading advantages in inference speed, consistency, dynamics, and semantic understanding.

Zhipu AI contributed to the "Beijing Major Open-Source Achievements" through its open-source foundation models. As China's earliest AI company to open-source large models, Zhipu AI has been committed to promoting AI accessibility. 2025 will be Zhipu AI's open-source year, with plans to release and open-source new large models (including foundation models, reasoning models, multimodal models, and agents).

08/

Five Qiming Venture Partners Portfolio Founders

Selected for China AI 100 Young Pioneers List

At this year's forum Future AI Pioneer Forum, DeepTech and MIT Technology Review officially released the China AI 100 Young Pioneers list. Fan Bao, co-founder and CTO of Shengshu Technology; Guohao Dai, co-founder and chief scientist of Infinigence AI; He Wang, founder and CTO of Galaxy Universal; Linfeng Zhang, founder and chief scientist of DeepWise; and Xiangyu Zhang, co-founder and chief scientist of StepFun, were selected.

The expert review process comprehensively evaluated candidates across technological innovation, industrial deployment, industry impact, interdisciplinary exploration, and social responsibility to select young pioneer representatives working in AI in China.

09/

Thorough Future Selected for 8th Zhongguancun International

Frontier Sci-Tech Competition AI Top 10

The finals of the 8th Zhongguancun International Frontier Sci-Tech Competition recently concluded, with Thorough Future selected for the AI Top 10 list. This honor not only demonstrates Thorough Future's leading strength in pathology AI but also reflects the profound value of artificial intelligence in the healthcare industry.

As a leading enterprise in pathology AI, Thorough Future has launched the discriminative pathology AI diagnostic product Thorough Insights. This product relies on the Thorough Brain pathology large model, focusing on precise diagnosis of high-incidence tumor organs, efficiently identifying malignant tumors and performing subtyping, while providing intelligent diagnostic support for benign lesions. The system precisely marks cancerous regions through heat maps and automatically generates structured pathology diagnostic reports.

Past Highlights

Qiming View | Qiming Venture Partners' Duane Kuang: AI is a Key Focus of Tech Investment, Chinese Engineering Design Product Capabilities Radiate Globally

Qiming Honors | Qiming Venture Partners and Investors Selected for 2023-2024 Tech and Manufacturing Investment Competitiveness Research Cases

Qiming Honors | Qiming Venture Partners and Portfolio Companies Selected for 2023-2024 China Tech Industry Investment List

Founded in 2006, Qiming Venture Partners currently manages 11 USD funds and 7 RMB funds, with total committed capital of $9.5 billion. Since its establishment, it has focused on investing in outstanding early- and growth-stage companies in Technology and Consumer (T&C) and Healthcare.

To date, Qiming Venture Partners has invested in over 580 high-growth innovative companies, of which more than 210 have listed on the New York Stock Exchange, NASDAQ, Hong Kong Exchanges and Clearing, Shanghai Stock Exchange, and Shenzhen Stock Exchange, or exited through M&A and other means. Over 80 portfolio companies have become recognized unicorns or super-unicorns.

Many Qiming Venture Partners portfolio companies have grown into the most influential companies in their respective fields, including Xiaomi (01810.HK), Meituan (03690.HK), Bilibili (NASDAQ:BILI, 09626.HK), Zhihu (NYSE:ZH, 02390.HK), Roborock (688169.SH), UBTECH (09880.HK), WeRide (NASDAQ:WRD), Gan & Lee Pharmaceuticals (603087.SH), Tigermed (300347.SZ, 03347.HK), Zai Lab (NASDAQ:ZLAB, 09688.HK), CanSino Biologics (688185.SH, 06185.HK), Schrödinger (NASDAQ:SDGR), MicroPort EP MedTech (688617.SH), Sanyou Medical (688085.SH), Amoy Diagnostics (300685.SZ), Berry Genomics (000710.SZ), GenScript ProBio (688520.SH), Yuanxin Technology, ClinChoice, Belief BioMed, Biren Technology, and others.