MoleculeMind
Category: AI in Biotech / Drug Discovery
AI protein design platform company developing a full-stack AI-powered protein engineering operating system (MoleculeOS) for drug discovery and biomanufacturing MoleculeMind was founded in 2022. The company is led by Jinbo Xu (许锦波). Based in Beijing, China. Team size: 101-500. Total funding raised: $100.0M. Latest round: Series A. Key investors include Sequoia Capital China (红杉中国), Cathay Biotech (凯赛生物), Cherami Investment, Shenzhen Capital Group (深创投), Blue Bridge Capital (蓝桥资本), Pudong Venture Capital (浦东创投), COFCO Emerging Industry Fund (中粮新兴产业基金), Oriental Fortune Capital (东方富海), Fosun Venture Capital (复星创富), Baidu Ventures, Lenovo Capital (联想创投), SenseTime Guoxiang Capital.
AMW Analysis
MoleculeMind develops an AI-powered protein engineering platform called MoleculeOS, designed for drug discovery and biomanufacturing. The company aims to transform protein engineering from trial-and-error into programmable bioengineering, enabling de novo design of antibodies, enzymes, and industrial proteins. The company has raised over $100 million across a Series A round, with investors including Lanqiao Capital, Pudong Venture Capital, COFCO Emerging Industry Fund, Fortune Capital (Fosun), and existing backers Cathay Biotech and Xinwan Capital.
In July 2026, MoleculeMind publicly launched MoleculeOS (MOS) as an AI-native operating system for biological R&D at the Shanghai State Capital Frontier Forum. The system integrates the company's proprietary model stack, including the NewOrigin (Darwin) multimodal protein model. This launch, combined with the recent funding, signals a platform shift from a protein design tool toward an infrastructure layer for broader biology R&D. The news flow indicates a strategic transition from a focused protein design platform to a full-stack operating system for biological research, supported by significant capital from a diverse investor base.
AMW analysis, generated from 2 tracked news signals.
- Founded
- 2022
- Headquarters
- Beijing, China
- Team size
- 101-500
- Total funding
- $100.0M
Value proposition
MoleculeMind's MoleculeOS platform transforms protein engineering from random trial-and-error into programmable bioengineering, enabling de novo design of antibodies, enzymes, and industrial proteins with dramatically higher hit rates and lower costs
Products and solutions
MoleculeOS (MOS) - AI-native bioengineering infrastructure platform, NewOrigin (达尔文) - multimodal protein foundation model, MMFold - high-precision protein structure prediction model (outperforms AlphaFold 3 on antibody-antigen interfaces per FoldBench), MMDesign - AI-powered de novo biologics design platform for nanobody/antibody discovery
Unique value
World's first full-stack AI protein operating system (MoleculeOS) powered by the NewOrigin multimodal protein foundation model, achieving >90% target hit rate in de novo antibody design across 12 clinical targets, with only 14-50 AI-generated candidates needed for wet-lab validation vs. billions in traditional screening
Target customer
Pharmaceutical and biotech companies (drug discovery R&D); Industrial biomanufacturing companies (enzyme engineering); Synthetic biology companies; Food and chemical material companies
Industries served
Biopharmaceuticals (antibody/nanobody drug discovery), Industrial biotechnology (enzyme engineering), Green biomanufacturing, Food technology, Chemical materials
Technology advantage
Proprietary NewOrigin multimodal protein foundation model; MMFold structure prediction surpassing AlphaFold 3 on complex antibody-antigen interfaces; MMDesign achieving >90% target hit rate with minimal wet-lab validation; Full-stack platform (MoleculeOS) covering design-prediction-optimization; Data flywheel from real industrial projects generating proprietary closed-loop data; Team led by AI protein folding pioneer Jinbo Xu (RaptorX-Contact inventor)
How they differentiate
Unlike competitors focused on single-model protein structure prediction, MoleculeMind has built a full-stack AI operating system (MoleculeOS) that covers the entire protein engineering workflow from design to wet-lab validation. Their MMFold model has demonstrated superior accuracy over AlphaFold 3 on antibody-antigen interfaces, and their MMDesign platform achieved >90% target hit rate with only 14-50 candidates per target — a step-change improvement over traditional billion-molecule screening approaches
Main competitors
Helixon (华深智药), TianRang XLab (天壤XLab), XtalPi (晶泰科技), EvolutionaryScale, Absci, Generate Biomedicines
Key partnerships
Cathay Biotech (凯赛生物) - synthetic biology leader and strategic investor, Sepax Technologies (赛分科技) - chromatography/purification technology collaboration, Partnerships with top global pharmaceutical companies for industrial enzyme optimization, Collaboration with Tsinghua University AIR
Notable customers
Global top-tier multinational pharmaceutical companies, US top-tier biotech venture capital-backed companies, China's innovative drug first-tier enterprises, Domestic synthetic biology leader (Cathay Biotech), Food industry listed companies, Global leading chemical materials group
Major milestones
2022-01: Company founded by Prof. Jinbo Xu, 2022-04: $5M+ Angel round led by Sequoia China, 2023-02: CNY 100M+ Strategic round led by Cathay Biotech, 2024-09: Series A co-led by Cherami Investment and Shenzhen Capital Group, 2025-2026: MMDesign platform achieves >90% target hit rate across 12 clinical targets, MMFold surpasses AlphaFold 3 on antibody-antigen interface prediction, 2026-06: Series A extension of $100M+ from diverse investor syndicate
Growth metrics
Achieved >90% target hit rate in de novo antibody design across 12 clinical targets; Solved pH-sensitive antibody design in 2 months (60x affinity difference); Increased expression level 400x for a poorly expressed pipeline asset; Multiple repeat customers with expanding project scope
Market positioning
Positioned as the leading AI protein engineering infrastructure platform in China, with global competitive standing — described as "AI-era Merck" and "AI protein industry infrastructure definer" by Chinese media. Competing globally against both Chinese peers (Helixon, TianRang XLab, XtalPi) and international players (EvolutionaryScale, Absci, Generate Biomedicines, Cradle)
Geographic focus
China (primary), with R&D presence in both China and the US; targeting global pharmaceutical and biomanufacturing markets
Patents and IP
Proprietary NewOrigin multimodal protein foundation model; MMFold protein structure prediction model; MMDesign de novo biologics design platform; MoleculeOS full-stack AI protein engineering platform (all self-developed IP)
About Jinbo Xu (许锦波)
Tenured Full Professor at Toyota Technological Institute at Chicago (TTIC, affiliate of University of Chicago); Distinguished Visiting Professor at Tsinghua University AIR; Inventor of RaptorX-Contact (first deep learning method for protein structure prediction, 2016); PhD in Computer Science
Latest news about MoleculeMind
More AI in Biotech / Drug Discovery companies
Official website: https://moleculemind.com/
