
Beijing-based BioGeometry closes Series A+ extension with $100 million funding led by China Life Capital, accelerating development of AI models GeoFlow and GeoCell for de novo drug design and cell-level prediction.
On average, it takes about 10-15 years to produce a new drug. Artificial intelligence could rewrite how new drugs are discovered as BioGeometry raises $100 million.
Artificial Intelligence (AI) could revolutionize the process of creating new drugs after BioGeometry raised $100 million.
The Beijing-based startup completed a Series A+ extension round, taking their total Series A funding to over $100 million. China Life Capital led the latest round.
With less than 4 years under its belt, BioGeometry is creating AI models capable of identifying new drug leads in weeks, rather than years. This could dramatically improve the economics of drug discovery.
BioGeometry Raised $100 Million Series A Round Led By China Life Capital
The AI-native biopharma company BioGeometry revealed their $100 million round was led by China Life Capital. Existing shareholder Shanghai Biomedical Industry Fund who co-led the initial round increased their investment. Other participants included China Development Venture Capital, Reliance Investment and Guofang Venture Capital.
Dacheng Wealth Capital also increased its investment. The round was exclusively advised by Index Capital. China Life Capital described BioGeometry as a "benchmark enterprise" of China's AI-native biotech industry.
This follows a pre-A round in 2024 where the company raised funds led by Jiangmen Ventures.
GeoFlow & GeoCell AI Drug Design Models
How BioGeometry's AI Model Works
BioGeometry uses two scale AI engines. The first is GeoFlow which focuses on molecules.
This is an all-atom model which can predict and design novel biomolecules. The initial version launched back in June 2024 performed as well as AlphaFold 3 in predicting protein complexes.
From there they continued to develop. In April 2025 they launched their V2 model which became the first Chinese AI model for de novo antibody design. In October 2025 they released their V3 model.
They added a generate-evaluate-optimize loop to their system, achieving an average hit rate of 20% across over 20 targets. They were able to identify leads in just 3 weeks. A larger V4 model is coming soon.
The second is GeoCell which looks at the cellular level. This model can help teams identify potential targets, diseases and appropriate patients by predicting cellular responses to genes, drugs or combinations.
Both models are integrated with an AI agent which connects them with their automated lab. All experiments will teach the system new things.
BioGeometry's Plans
BioGeometry will invest the new capital into improving their GeoFlow and GeoCell models. They will also grow their high throughput automated labs. Their pipeline will also be accelerated with their own drug projects already having dozens of projects in the pipeline.
A number of bispecific antibody programs are due to enter IND filing in 2027. Synthetic biology projects could start generating revenue sooner. Over 5 projects have completed their pilot testing.
One standout project is the natural borneol compound used in traditional medicine. BioGeometry says they are the first to create it using biosynthesis, reducing the cost by around 80% compared to extracting it from plants. They expect to generate large revenues in 2027.
Drug discovery is a long, expensive process with most candidates failing. Using AI gives much better chances of success by designing smarter molecules. BioGeometry are confident they can achieve this.
They are generating revenues from their synthetic biology projects too. This gives investors a unique combination of a moonshot and immediate returns. Overall,
China's AI biotech