Beyang Therapeutics has raised nearly $30 million in an oversubscribed Series A round, giving the China-based biotech fresh capital to advance an AI-driven drug discovery platform and two differentiated therapeutic programs. The financing comes as pharmaceutical companies increasingly look beyond generative AI experimentation toward computational systems that can produce experimentally validated drug candidates.

Beyang Therapeutics is betting that the next phase of AI drug discovery will be less about generating promising molecules on a screen and more about integrating computational design with the experimental work required to turn those molecules into medicines.

The Shanghai-based company said it has closed an oversubscribed Series A financing of nearly $30 million, co-led by Legend Capital and Shanghai Healthcare Capital. China Medical System Holdings, Fenglei Capital, Shanghai Sci-Tech Innovation Center Capital and Harbor & Canton Capital also participated, alongside existing investors Root Venture Partners and YuanBio Venture Capital.

Beyang plans to use the funding to develop its ExCEED AI and molecular-design platform, advance pipeline programs through clinical and preclinical development, and expand its international research and business-development operations.

Founded in 2021, Beyang describes itself as an AI-driven innovative drug R&D company. Its technology combines proprietary datasets, artificial intelligence and computational simulation across molecular design, druggability optimization and biological evaluation.

The strategic question for the company is whether that integration can translate into a repeatable advantage in discovering molecules that are both novel and therapeutically viable.

Moving beyond AI-assisted molecule generation

AI drug discovery has attracted substantial investment from pharmaceutical companies and biotech startups, but the field faces a persistent challenge: identifying molecules that look compelling computationally is not the same as producing drug candidates that survive laboratory and clinical testing.

Beyang says ExCEED is designed to address that gap by covering multiple stages of the discovery process.

A key claim is its ability to generate new molecular scaffolds, rather than simply modifying existing chemical cores. In drug development, that distinction can matter because incremental variations of established structures can face crowded intellectual-property landscapes, biological resistance or limited differentiation.

Beyang says the platform ranked eighth globally among more than 350 participants in the OpenADMET blind prediction challenge, which tested computational predictions related to absorption, distribution, metabolism, excretion and toxicity characteristics.

That result provides an external data point for evaluating the platform, although a competition ranking is not equivalent to clinical validation. The more consequential test will be whether candidates generated or optimized through ExCEED can demonstrate durable advantages in human studies.

Two programs give the platform a clinical test

Beyang’s pipeline currently includes programs in ophthalmology and oncology.

BT01001 is a small-molecule eye drop being developed for retinal diseases. The company says preclinical head-to-head studies showed ocular-tissue exposure comparable to its stated target benchmarks and efficacy comparable to intravitreal aflibercept in those studies.

The program has completed Phase I single-ascending-dose and multiple-ascending-dose studies in healthy volunteers, according to the company, with positive safety and tolerability findings. Beyang is preparing for first-in-patient trials.

The commercial and clinical proposition is significant if the approach succeeds. Many retinal diseases are currently treated with injections delivered directly into the eye. A sufficiently effective topical therapy could potentially reduce the burden associated with invasive administration.

But that is also where development risk becomes more pronounced. Ocular exposure, tissue penetration and efficacy demonstrated in animal or early-stage studies do not guarantee that an eye-drop formulation will achieve the required therapeutic concentration in patients.

The company’s second lead program, BT01002, targets acute leukemia associated with KMT2A rearrangements or NPM1 mutations. It is a second-generation Menin inhibitor designed to address resistance mechanisms associated with earlier-generation compounds.

Beyang says BT01002 is currently in IND-enabling studies and that its molecular design addresses known resistance mutations. If validated clinically, the program could enter a competitive field in which the ability to maintain activity against resistant disease is a central differentiator.

AI drug discovery enters a more demanding phase

The financing reflects a broader transition in the AI drug discovery market.

Early AI-biotech narratives often centered on whether machine learning could predict molecular properties or identify potential drug candidates faster than conventional approaches. The industry’s next test is more practical: can integrated AI platforms consistently produce differentiated assets that reach clinical milestones?

Large technology companies are also pushing deeper into the sector. NVIDIA is supplying computing infrastructure and drug-discovery software ecosystems, while Microsoft has developed AI capabilities for life sciences. Pharmaceutical companies and specialist drug-discovery firms are increasingly combining machine learning with structural biology, molecular simulation and laboratory automation.

That makes platform differentiation increasingly important.

Beyang’s positioning is closer to a vertically integrated drug-discovery model than a standalone AI model provider. Its argument is that combining computational chemistry, biological evaluation and proprietary data can create a feedback loop in which experimental results improve subsequent molecular design.

For enterprise pharmaceutical teams, that architecture is potentially more relevant than a generic AI model. Drug companies ultimately need reproducibility, experimental evidence, intellectual-property protection and a path from discovery to regulatory submission.

International expansion signals a partnering strategy

Beyang also says it intends to expand international R&D and business-development teams while seeking external collaborations.

That could become an important part of the company’s next phase. AI drug discovery platforms can generate value through internal pipelines, but partnerships with pharmaceutical companies can provide additional validation, development expertise and access to global clinical infrastructure.

The funding therefore gives Beyang two related opportunities: advance its own assets while demonstrating that ExCEED can support programs outside its internal portfolio.

The company will need to show that its technology can repeatedly deliver differentiated molecules—not simply one successful candidate.

For the broader AI-biotech sector, that distinction is becoming critical. The industry’s winners are unlikely to be determined by who can produce the most molecules with AI. They will be determined by who can turn computational predictions into safer, differentiated medicines with credible clinical and commercial paths.

Market Landscape

AI-enabled drug discovery is moving toward vertical integration, bringing together generative chemistry, molecular simulation, biological testing, proprietary datasets and laboratory validation.

That shift creates opportunities for companies such as Beyang, but also intensifies competition. NVIDIA is building infrastructure and software for computational drug discovery; Microsoft and other hyperscalers are embedding AI into life-sciences workflows; while specialist companies and pharmaceutical R&D organizations are developing proprietary models and datasets.

The strategic advantage of platforms such as ExCEED will ultimately depend on their ability to create a measurable improvement in the drug-development funnel: better candidates, fewer failed experiments, faster optimization or stronger clinical differentiation.

For pharmaceutical enterprise teams evaluating AI drug-discovery partnerships, three factors matter particularly: validation data, proprietary datasets and integration with experimental workflows. A compelling AI model alone is unlikely to be sufficient.

Beyang’s Series A provides the capital to test that proposition through BT01001 and BT01002 while expanding its platform and external collaboration strategy.

Top Insights

• Beyang raised nearly $30 million to scale ExCEED, an AI drug-discovery platform combining molecular design, simulation and biological evaluation for pharmaceutical R&D teams.
• The company is advancing BT01001 and BT01002, targeting retinal disease and resistant acute leukemia while testing whether AI-generated molecules can deliver clinical differentiation.
• ExCEED ranked eighth among more than 350 participants in an OpenADMET blind prediction challenge, providing external evidence for its computational modeling capabilities.
• Beyang’s strategy reflects AI drug discovery’s shift toward vertically integrated platforms linking generative chemistry, proprietary data, simulation and experimental validation.

Get in touch with our fintech expert