XtalPi Holdings Announces 2025 Interim Results

USA - English

• APAC - English
• APAC - Traditional Chinese

SHENZHEN, China, Aug. 27, 2025 /PRNewswire/ --

Key Highlights from the Financial Report:

• Revenue surged 404% year over year to RMB 517 million in the first half of 2025.
• Adjusted net profit reached RMB 141 million, marking the company's first-ever half-year profitability.
• XtalPi was added to the MSCI China Small Cap Index, underscoring strong recognition of its investment value by the international capital markets.
• Significant breakthroughs and milestones were achieved across multiple business lines.

Drug Discovery Solutions: The Deep Integration of Advanced AI and Robotics Driving New Growth for Our Business

During the reporting period, revenue from our Drug Discovery Solutions business soared by 615.2% year-over-year, from RMB 60.85 million in H1 2024 to RMB 435.2 million. This growth was driven primarily by substantial upfront payments from major collaborations and multiple pipelines hitting key milestones.

1)   Landmark Collaboration with Professor Gregory Verdine

In late June, we signed a strategic collaboration with renowned biopharmaceutical scientist Professor Gregory Verdine. The agreement provides for an initial payment of US$51 million for the first phase, with an additional US$49 million in potential payments, and variable payments of up to US$ 5.89 billion subject to reaching regulatory and commercial milestones, along with potential royalties based on a single-digit percentage of the annual net sales of the products. This collaboration represents a key achievement in our global strategy and further validates the industry-leading capabilities of our AI+robotics drug discovery platform.

2)   Multiple Pipelines Achieve Breakthrough Milestones

• Chronic Disease Collaboration Ahead of Schedule: All three of our projects in collaboration with a leading biopharmaceutical company have successfully reached key milestones. Notably, a chronic disease program launched in 2024 achieved its PCC milestone a full quarter earlier than planned, highlighting the power of XtalPi's fully integrated AI-computational-experimental platform in accelerating drug discovery.
• PRMT5 Inhibitor PE-0260 Advanced to IND: In close collaboration with PharmaEngine, we successfully developed the first-in-class PRMT5 inhibitor PE-0260. PharmaEngine has now completed IND application and initiated clinical trials.
• Prix Galien Award Nomination: The world's first targeted therapy for diffuse gastric cancer, co-developed with incubated company Signet Therapeutics, was nominated for the prestigious Prix Galien Award for Best Biotechnology Product, often hailed as the "Nobel Prize of the pharmaceutical industry". This was the only Asian pipeline nominated this year and only the fourth Chinese company to be nominated in the award's 50+ year history.
• Antibody Business Growth: Thanks to high-quality deliveries of previously placed orders, in the first half of 2025, we recorded antibody business revenue nearly double that of full-year 2024. In this area, we now have partnerships ongoing with several globally renowned pharmaceutical companies, with our role evolving from a service provider to platform-based co-developer, significantly increasing customer lifetime value.
• Breakthroughs at XtalPi-incubated Leman Biotech: (1) Leman's metabolically enhanced CD19 CAR-T therapy successfully treated multiple systemic lupus erythematosus (SLE) patients, while its solid tumor CAR-T program advanced rapidly toward investigator-initiated trial (IIT) readiness. (2) Leveraging AI-driven optimization and high-throughput experiments, we improved the binding affinity and immune activity of its core metabolic enhancers, enabling the project to advance into IIT preparation in the first half of 2025.

3)   Enhancing AI Capabilities to Create a Drug Discovery Superintelligence

• We have developed over 200 AI models spanning molecular generation, free energy perturbation (FEP), property prediction, and crystal structure prediction. In the first half of 2025, building on our earlier strategic partnership with Pfizer, we continued to iterate and optimize these models while deepening our collaboration to co-develop a next-generation, AI-driven molecular simulation platform. Additionally, we are actively expanding the platform's capabilities across diverse applications, including small molecules, peptides, and antibodies, to comprehensively empower Pfizer's drug discovery efforts.
• We launched a new Ultra mode for our AI antibody development platform XtalFold^®, improving antigen-antibody complex prediction accuracy by approximately 10 percentage points. In July, our self-developed AI antibody drug structure modeling platform XtalFold^® was honored one of the "Top 10 AI Innovations (Technologies/Products)" at the World Artificial Intelligence Conference (WAIC).
• We initiated the development of a Molecular Glue platform built around a Design — Make — Test — Analyze (DMTA) closed-loop framework in new drug development. Our core Molecular Glue team is now in place and has established initial collaboration with key opinion leaders from both industry and academia. Meanwhile, the team is conducting in-depth discussions with multiple pharmaceutical companies on specific targets and screening strategies. Initial feedback has been positive, and partnership intentions are steadily progressing.
• We launched PepiX™, a global-leading peptide and protein R&D platform that integrates generative AI molecular design, automated synthesis, trillion-scale peptide libraries, and high-throughput screening. Early results include:
  • In AI-driven cancer vaccine development, candidate molecules designed by the platform have demonstrated strong immunogenicity in preclinical animal models and are poised to enter clinical trials
  • Two independently developed oral peptide health products for blood sugar regulation and fat reduction have completed efficacy and toxicology evaluations and are now entering the regulatory approval process.
  • A successful first-in-class renal cancer target development project with National Cancer Centre Singapore (NCCS) is progressing to downstream efficacy validation.

We have also built a library of over 2,000 non-natural amino acid monomers, fully parameterized for computational modeling, enhancing molecular design and exploration.

• At the AAAI (Association for the Advancement of Artificial Intelligence) 2025, we presented our breakthrough research mRNA2vec, an AI-driven technology that optimizes mRNA sequence design, enhancing expression levels and stability while addressing the high costs and lengthy timelines of traditional screening methods, establishing a new paradigm for intelligent mRNA designs.

Intelligent Robotics Solutions: AI and Robotics Driving a Transformation in Chemistry

Revenue from our Intelligent Robotics Solutions business grew 95.9% year-over-year to RMB 81.9 million in the first half of 2025, fueled by the rapid growth from our automated chemical synthesis services and our XtalPi R&D solutions.

1)   AI + Robotics Reshape the R&D Paradigm, Building Vertical Artificial Superintelligence

We have built a unique flywheel system integrating high-throughput experiments, high-quality data, and advanced AI models. Our robotics lab operates 24/7, generating vast volumes of high-quality data to power vertical AI models. This paradigm is transforming traditional R&D approaches, accelerating breakthroughs in both drug and material development.

Our AI models and robotic laboratory have been widely applied across the field of chemistry, significantly expanding the accessible chemical space, overcoming critical bottlenecks in drug and new material molecule development, and effectively accelerating delivery timelines. During the reporting period, we also made significant progress in several key technologies:

• The Future of Chemistry: Innovation Breakthroughs Powered by Models and Data
  • In the chemical reaction stage, leveraging extensive experimental data from our robotics lab and knowledge graphs obtained via the LLM-powered literature search tool PatSight, we have successfully developed AI models for synthetic feasibility prediction and reaction condition recommendation covering common reaction types.
  • In practical applications, these models have boosted the success rate of predicting chemical reaction success rates to over 85%.
  • In the separation and purification phase, we developed reaction condition recommendation and yield prediction models tailored for condition screening, which have been successfully applied in multiple client projects. In two of these projects, yields exceeded 50%.
  • Additionally, we have completed the acquisition of Liverpool ChiroChem (LCC), a company with deep expertise in chiral chemistry, strengthening our capabilities in chemical space exploration. LCC boasts a world-leading library of chiral molecules and a high-quality database, offering significant advantages in molecular quality and diversity and providing strong data support for AI-driven exploration of the chiral chemical space.
• Breakthrough in Embodied Intelligence: "NeoDispenser" Solves Key Challenges in Robotic Laboratories

In the first half of 2025, we achieved significant breakthroughs in robotic laboratories, notably advancing embodied intelligence technology. We successfully developed a new generation of modular robotic systems that integrate advanced visual perception, AI, and robotic technologies to tackle some of the most challenging problems in chemical research.

For example, our next-generation modular robot features our patented "NeoDispenser," a flexible gripper that successfully mimics human hand operations with exceptional dexterity, earning recognition from top global clients. This design addresses three major industry challenges: (1) precise handling of trace powders; (2) overcoming traditional solutions' poor adaptability to different powders; (3) eliminating intermediate containers to reduce costly material waste.

• Multi-Agent Systems Driving a Revolution in Chemistry: From Massive Chemical Space Exploration to Fully Autonomous Decision-Making

In the first half of 2025, we successfully built a cost-effective, high-performance data platform capable of predicting synthetic feasibility across hundreds of billions of chemical compounds.

• LLM-powered drug R&D data extraction — PatSight Patent Mining Platform Accelerates Structure Analysis

Our PatSight patent mining platform, jointly developed and iterated with the IDEA Research Institute, has achieved breakthrough progress powered by LLMs and deep neural network technologies. The platform can efficiently extract core data from literature and patents, such as molecular structures, activities, pharmacological effects, and reactions, with an accuracy rate of up to 97%, significantly enhancing the efficiency of structure analysis and data acquisition. PatSight overcomes the inefficiencies of traditional manual processing, providing high-quality public big data support for AI4S models in specialized domains. In the first half of 2025, the platform secured multiple procurement and collaboration agreements with domestic and multinational pharmaceutical companies.

• By consistently upgrading the materials research platform and exploring applications across multiple material domains, we have continued to expand our technological applications across new chemical materials, including adhesives, coatings, electrolytes, catalysts, and advanced carbon materials. We have also secured orders from several leading companies in the materials sector, with overall project progress advancing smoothly.

2)   XtalPi R&D Solutions Achieve Global Commercial Breakthroughs

As our technological capabilities continue to grow, our XtalPi R&D solutions have gained strong customer recognition and are rapidly expanding into international markets. During the reporting period:

• We signed a multi-million US Dollar agreement with South Korea-based JW Pharmaceutical, one of Asia's leading pharmaceutical companies. Under the deal, XtalPi will deliver a comprehensive solution that includes a high-throughput automated synthesis workstation, an AI-driven reaction condition optimization system, and an intelligent analysis platform. Featuring Bayesian optimization algorithms, digital experiment management, and advanced automated lab technologies, this integrated solution will significantly boost synthesis success rates and repeatability while shortening drug development cycles. The collaboration will provide JW with a robust technological foundation to accelerate breakthrough pipeline development in oncology, metabolic diseases, and other key areas, fast-tracking more globally competitive drug candidates into clinical trials.
• We signed a multi-million US Dollar cooperation agreement with Liangzhu Laboratory. Under this partnership, XtalPi and Liangzhu will collaborate on setting up a robotics-powered biomaterials laboratory, with XtalPi providing a smart robotics platform spanning their entire workflow from synthesis to testing.
• We entered into a key partnership with Roche, under which XtalPi will integrate robotics, transport and barcode scanning modules, stacker cranes, and automated pallet and tube handling systems to enable automated sorting, high-throughput storage, and precise retrieval of molecular building blocks.

PROSPECTS AND OUTLOOK

We believe the future of scientific discovery lies in the seamless fusion of artificial intelligence, embodied intelligence, and human insight. We are boldly tackling challenges to unlock new frontiers, transforming technological barriers into breakthroughs and visionary ideas into impactful solutions. By building superintelligent systems for life and materials sciences, we seek to reshape the future of human civilization through a pioneering new research paradigm.

About XtalPi

XtalPi Holdings Limited (XtalPi, 2228.HK) was founded in 2015 by three physicists from the Massachusetts Institute of Technology (MIT). It is an innovative R&D platform powered by quantum physics, artificial intelligence, and robotics. By integrating first-principles calculations, AI algorithms, high-performance cloud computing, and standardized automation systems, XtalPi provides digital and intelligent R&D solutions for companies in the pharmaceutical, materials science, agricultural technology, energy, new chemicals, and cosmetics industries.

Investor Relations Contact:

In China:

XtalPi Holdings Limited
Investor Relations
Email: [email protected]

Piacente Financial Communications
Tel: +86-10-6508-0677
Email: [email protected]

In the United States:

Piacente Financial Communications
Brandi Piacente
Tel: +1-212-481-2050
Email: [email protected]

SOURCE XtalPi Inc.