Yunqi Capital Invests in Baiyao Technology: AI Virtual Cells Bring Life System Changes into the Computable Era | Yunqi Partners

AI is being used to understand and predict how living systems change under different interventions, providing new foundational tools for drug discovery, cell therapy, and disease research.

Recently, the Yunqi-Shanghai Jiao Tong University AI Angel Fund participated in a new funding round of tens of millions of RMB for Baiyao Technology. Founded in 2023, the AI Virtual Cell (AIVC) platform company has been building foundational large models for AI virtual cells, aiming to simulate and predict cellular dynamic responses to perturbations such as drugs, gene editing, and environmental changes through massive single-cell multi-omics data.

Join this issue's "Yunqi Partner" to learn how Baiyao Technology is exploring next-generation infrastructure for AI + life sciences through a "model-data-scenario" closed loop.

Yunqi Capital's Perspective:

We continue to track breakthrough opportunities in AI for Science across various domains. AI virtual cell models represent a critically important technological breakthrough in the life sciences in recent years — in principle, transforming life science development into a computable and predictable digital paradigm. Baiyao's model capabilities are globally leading, with both general-purpose and domain-specific models already achieving commercial deployment with top-tier global clients. We look forward to the company's continued breakthroughs in the life sciences field.

The following content is reprinted from 36Kr

AI Virtual Cell (AIVC) platform company Baiyao Technology recently completed a new funding round of tens of millions of RMB. This round was led by Guoxin Venture Capital Fund, under the state-owned capital operation platform China Reform Holdings. The proceeds will be mainly used for next-generation virtual cell algorithm model iteration, exclusive data collection platform construction, and accelerated industrialization.

Baiyao Technology's founding team draws on research resources from the Chinese Academy of Sciences' Institute of Zoology and the Beijing Institute of Stem Cell and Regenerative Medicine. Since 2023, they have been building foundational large models for AI virtual cells to enable precise analysis and prediction of cellular behavior and state changes, while continuously advancing AIVC model upgrades and industrialization efforts.

Since 2026, funding enthusiasm in the AI drug discovery track has remained high, and industry attention has gradually shifted from early single-point breakthroughs in molecular screening toward deeper cell-level life system modeling and dynamic simulation. Unlike AI drug discovery focused on molecular conformation optimization, AIVC's value spans the entire drug development lifecycle: from disease mechanism analysis, target discovery, candidate drug virtual screening, preclinical efficacy and toxicity prediction, to clinical patient stratification. Its core breakthrough lies in elevating AI modeling from the "molecular" level to the "cellular" level, directly simulating living systems' dynamic responses to external interventions.

AIVC is trained on massive single-cell multi-omics data, learning cellular state transition patterns, gene regulatory networks, and cross-scale causal relationships to build foundation models capable of representing cellular behavior and fate evolution. Researchers need only input initial cellular states (such as induced pluripotent stem cells, immune cells, or tumor cells) and specific perturbation conditions (such as drug action, gene editing, or environmental changes), and the model can predict future state transitions, including proliferation, differentiation, senescence, apoptosis, and disease progression.

This capability allows researchers to conduct large-scale virtual experiments in digital environments first, prioritizing research hypotheses with the highest probability of success, then validating through wet-lab experiments. Thus, the R&D process gradually shifts from traditional "experiment-driven" to "model-driven" approaches, potentially significantly reducing R&D costs, shortening development cycles, and improving overall drug development success rates.

From an industry development perspective, AIVC is evolving from a pure R&D tool into new infrastructure for the life sciences. Just as large language models reshaped information processing, AIVC is poised to become a core platform for understanding, predicting, and designing living systems, providing foundational capabilities for next-generation drug discovery, cell therapy, and synthetic biology. Accordingly, ARK Invest stated in its Big Ideas report that virtual cells would be "one of the most disruptive areas in future AI + life sciences."

Currently, primary market capital flows are gradually validating this judgment. Baiyao Technology's founding team stated that this funding round, recognized by a state-owned capital operation platform and market-oriented technology and healthcare investors, will help the company rapidly complete next-generation foundation and perturbation models, deploy proprietary data platform construction and exclusive data partnerships. While ensuring core AIVC capabilities, they will pioneer industrial application scenarios, deploying advanced cell therapy and "AIDD plus AIVC," accelerating the "model-data-scenario" industry chain closed loop.

It is reported that as early as September 2023, the founding team released GeneCompass, a knowledge-enhanced multi-species cell large model, published as a cover article in Cell Research. This model is a general-purpose virtual cell foundation model with over 100 million single-cell data points and embedded knowledge. In 2025, the team further upgraded and released the CellGraphCompass model, systematically applying graph structure algorithms and prior knowledge to virtual cell modeling. That same year, perturbation models built on the foundation model helped the founding team achieve first place globally in the preliminary round and second place overall in the finals of the global virtual cell challenge.

Additionally, Baiyao Technology has accumulated substantial data assets. Beyond relying on the founding team's four years of accumulated global large-scale AI-ready public datasets for foundation model training, the team will soon deploy proprietary data technology to collect high-quality temporally continuous perturbation datasets suited for AIVC model training, and launch large-scale clinical data collaboration projects for proprietary dataset generation.

On the application front, Baiyao Technology has already advanced deployment with domestic and international partners in model deployment, joint cell therapy pipeline R&D, early target discovery, and clinical sample precision stratification scenarios.