MoleculeMind Secures Over $100 Million in Series A Funding as Investors Bet on AI-Driven Biological Infrastructure

This round of financing was participated in by institutions including Blue Bridge Capital, Pudong Venture Capital, COFCO Emerging Industries Fund, Oriental Fortune Capital, Fosun RZ Capital, Guofang Venture Capital, Sepax Technologies, Infinity Group, Caixin Capital, and Zhengding. Existing shareholders such as Cathay Biotech and Xinhang Capital also continued to increase their investments. The diversified investor lineup fully demonstrates the deep consensus and joint commitment of financial investment institutions, industrial capital, and strategic guiding capital toward MoleculeMind.

Amid capital markets’ clamor for ultimate “certainty” and AI for Science (AI4S) entering deep waters, this financing round sends a clear signal:The global competition in the AI protein field has fully transitioned from “single-model benchmarking” in laboratories to an industrial-level contest focused on “solving real industry pain points and achieving commercial viability.”Amid this reshuffling, MoleculeMind is establishing its rare status as the definer of global AI protein industry infrastructure, leveraging world-leading foundational technologies, verifiable industrial-grade achievements, and a rare cross-sector capital consensus.

Over the past few decades, both new drug development and industrial enzyme engineering have essentially been a “blind box screening” process heavily reliant on luck. Faced with candidate molecule libraries numbering in the hundreds of millions, pharmaceutical companies often need to spend more than ten years and billions of dollars on screening and validation.

“The second half of AI drug discovery must bid farewell to random trial-and-error.”Professor Xu Jinbo, founder of MoleculeMind, believes.

Xu Jinbo, the inventor of RaptorX-Contact, the world’s first effective AI algorithm for protein structure prediction, is a world-class scientist hailed as the “founder of AI protein folding.” Following the first paradigm shift in protein structure prediction, he is now leading the MoleculeMind team to drive another transformation:Propelling the R&D of biomacromolecules from "random trial-and-error" into a deterministic era of "programmable bioengineering."

While most players remain focused on structure prediction, MoleculeMind has extended its capabilities into the industrial validation phase of “de novo design,” delivering a track record of “created” molecules against real clinical targets.

Just recently, MoleculeMind’s AI-powered de novo design platform for biologics, MMDesign, demonstrated highly disruptive industrial value:In tests against 12 real-world clinical targets, a hit rate exceeding 90% was achieved, successfully tackling industry-recognized “hard nuts” such as GPCRs and TNFα.

More critically, it delivers an exponential increase in R&D cost-efficiency and speed. While traditional methods require screening hundreds of millions of candidates to identify a single antibody, MMDesign needs only wet-lab validation of 14 to 50 AI-generated candidate molecules to obtain novel nanobodies with high affinity and high expression levels. Notably, the antibody designed against the ultra-challenging target TNFα achieves picomolar-level affinity.

This high hit rate is underpinned by MoleculeMind’s proprietary high-precision structure prediction foundation model, MMFold. In the internationally authoritative FoldBench benchmark, MMFold demonstrated prediction accuracy surpassing all open-source models when faced with extremely complex antibody–antigen interfaces, achieving a substantial lead over mainstream domestic and international models such as Google’s AlphaFold 3 in high-precision structure prediction. This means thatChinese Team Achieves Key Transition from "Catching Up" to "Leading" in the Most Industrially Valuable Core Sector of AI-Driven Protein Research for the First Time

A single technological breakthrough is insufficient to support large-scale industrial application. The core competitive barrier of MoleculeMind lies in its transcendence of the positioning as a mere algorithmic tool, by constructing a suite ofAI-Native Bioengineering Infrastructure——MoleculeOS (abbreviated as MOS,Website: https://mos.moleculemind.com/）。

MOS, built on MoleculeMind’s self-developed NewOrigin (Darwin) multimodal protein foundation large model, deeply integrates AI with first-principles thinking, bridging the “AI-Powered Precision Design — Wet-Lab Small-Sample Validation — Continuous Model/Platform Evolution” engineering flywheel.

Leveraging MOS, MoleculeMind has achieved consecutive victories in high-barrier industrial scenarios that have stymied traditional experiments and general-purpose AI, with validation through wet-lab experiments:

In the challenging frontier of innovative drug development, MoleculeMind solved the long-standing problem of pH-sensitive antibody design—previously unaddressed by traditional high-throughput screening—in just two months, achieving a 60-fold difference in affinity across varying pH environments. For a pipeline candidate with severely inadequate expression, it boosted expression levels by over 400-fold, with monomer purity exceeding 90%, thereby directly “reviving” a highly commercially valuable asset.

In the field of green biomanufacturing, collaboration with industry partners on the directed evolution and engineering of super industrial enzymes has increased strain yields several-fold, opening up significant potential for cost reduction and efficiency improvement in industrial-scale production.

Currently, the MOS platform has undergone real-world project validation by strategic, industry-leading partners across multiple sectors, including innovative pharmaceuticals, food, and advanced chemical materials. These partners include top-ranked multinational pharmaceutical companies by global market capitalization, premier U.S. biopharmaceutical venture capital firms, China’s first-tier innovative drug developers, leading domestic synthetic biology enterprises, publicly listed food industry companies, and globally leading advanced chemical material groups.Multiple partners proactively initiated follow-up collaborations after the completion of the initial project, with some clients establishing continuous multi-project partnerships. The scope of cooperation has gradually expanded from single projects to multiple targets, diverse molecular formats, and various application scenarios. This fully demonstrates that MoleculeMind’s AI protein design platform possesses strong customer stickiness, cross-domain applicability, and sustained value creation capabilities, laying a solid foundation and exhibiting tremendous potential to become a foundational R&D platform in the life sciences sector.

Meanwhile, these real-world industrial projects, having undergone rigorous validation through wet-lab experiments, are continuously accumulating into high-quality proprietary closed-loop data, thereby feeding back into and strengthening the MOS platform.This data flywheel, driven by “real-world industrial implementation,” has built an insurmountable moat for the industry and enabled MoleculeMind to establish a unique engineering advantage in the global AI protein sector.。

Currently, global innovative drug development is at a critical juncture for crossing economic cycles, with capital and industrial resources accelerating their concentration toward foundational "hardcore infrastructure." The industry's application of AI technology has moved beyond merely satisfying research-grade structure prediction; it now demands engineered solutions that cover the entire "design-validation-production" chain. Only platforms possessing "foundational technology + engineering capabilities + an industrial closed loop" can thrive in this environment.

MoleculeMind’s cross-domain foundational platform attributes and proven customer stickiness have made it a scarce asset attracting collective investment from both capital markets and industry players.

Huang Bohao, Founder of Lanqiao Capital“MoleculeMind is one of the very few teams globally that truly possess original, foundational AI capabilities in protein science. Professor Xu Jinbo’s academic stature is irreplaceable, but even more remarkable is their ability to translate technological prowess into verifiable commercial milestones. This closed-loop capability—from ‘scientific breakthrough’ to ‘industrial delivery’—is extremely rare in the field of AI-driven drug discovery.”

Chen Huawei, Partner at Oriental Fortune Capital“It is believed that ‘AI4S is one of the most certain directions for technological revolution in the next decade. MoleculeMind not only possesses world-leading protein foundation models, but has also established payable and verifiable capabilities in high-frequency, essential scenarios for pharmaceutical companies, such as antigen-antibody complex prediction and conditionally activated protein design.’”

The head of investment at COFCO Emerging Industries Fund stated“Biomanufacturing is a future-oriented industry within the national strategic layout. MoleculeMind has deeply integrated AI into areas such as industrial enzyme design, biomaterial optimization, and product innovation for the food industry, demonstrating significant commercialization potential. COFCO Industrial Fund will continue to empower MoleculeMind’s industrial implementation in the biomanufacturing sector, leveraging the scenario-based and data advantages of related industrial companies in the biological field, thereby jointly promoting China’s leading development in the AI-driven bioeconomy.”

Huang Xueying, Chairman of Sepax TechnologiesEmphasizing Synergistic Value: “Chromatographic purification and protein engineering are tightly coupled, critical links in the biopharmaceutical industry chain. MoleculeMind’s leading capabilities in AI-driven protein design and engineering are highly complementary to Sepax Technologies’ technical advantages in separation and purification, as well as its scaled manufacturing capabilities and industrial resources. We look forward to jointly enhancing the overall efficiency of R&D and production for innovative biologics.”

As a long-standing shareholder with continuous capital injections,Dr. Yang Chen, President of Cathay BiotechMoleculeMind’s capabilities in AI-driven protein engineering have been fully validated through our industrial collaboration. From the directed evolution of industrial enzymes to significant improvements in microbial strain yields, MoleculeMind has demonstrated the substantial value of AI in biomanufacturing with tangible results. As an existing shareholder, we continue to increase our investment precisely because we recognize the certainty of the “AI + Biomanufacturing” pathway and MoleculeMind’s rare ability to translate technology into industrial efficiency.

MoleculeMind has established a multi-tiered business model encompassing platform licensing, joint R&D, and proprietary pipelines. In the area of platform licensing, MOS has opened access to leading pharmaceutical companies, embedding its “AI-driven precision design + small-scale wet lab validation” capabilities into partners’ R&D workflows. In joint R&D, MoleculeMind collaborates with industry partners to continuously deliver industrial-grade outcomes for challenging targets and complex protein engineering projects. Regarding its proprietary pipeline, MoleculeMind has also accumulated multiple early-stage assets with First-in-Class potential.

Professor Xu Jinbo stated: “"The underlying technological revolution in AI-driven protein science has completed theoretical validation, transitioning from the laboratory to large-scale industrial application, thereby converting silicon-based computational power into tangible material value."