After NVIDIA, Eli Lilly invests $2.75 billion in Chinese AI-driven pharmaceutical development

Ask AI · Why Eli Lilly Chose China’s AI Drug Discovery Companies to Accelerate Drug Development?

China’s AI drug discovery is becoming a major player in the global market

**By | **Ling Xin

Edited by | Wang Xiao

Photo/Visual China

Just past the weekend, China’s artificial intelligence (AI) drug discovery company I-s? (I-S?—not sure of the proper English name) announced that it had won a major deal from multinational pharmaceutical company Eli Lilly.

The agreement mainly centers on a new oral therapeutic drug, but neither side intends to disclose which specific therapeutic area the molecule in question targets.

Under the deal, I? will receive a $115 million upfront payment, and after milestone payments, the total transaction value could reach approximately $2.75 billion; in addition, there will be a revenue share.

This transaction is not only a “big deal,” but also triggered antitrust review by the U.S. Federal Trade Commission (FTC). Information about the review was formally updated on March 26, 2026. According to the FTC’s official release, the filing threshold for transaction values in 2026 is around $130 million.

For Eli Lilly, this deal is also a major move. Earlier this year, Eli Lilly collaborated with U.S. AI drug discovery company Nimbus Therapeutics on something similar, but both the upfront payment and total deal value were only half of this transaction. Eli Lilly’s previous main collaborator included global AI giants such as NVIDIA.

Besides Eli Lilly, Sanofi also has a deal worth more than $1 billion with I? . Compared with traditional “innovative drug” pharmaceutical companies, China’s AI drug discovery firms seem to be crossing the “low-margin, high-volume” phase faster and have started becoming important participants in the global market, and their pricing is not cheap either.

“We have good value for money, but not cheap.” I? co-CEO and Chief Scientific Officer Ren Feng told Caijing on March 30, saying that compared with peers in the AI drug discovery field, their prices are still relatively high.

Why is Eli Lilly willing to put down more than $100 million upfront to bet on Chinese companies? Through the underlying reasons, you can see how global AI drug discovery is heading forward.

No.1

Why Partner with Chinese AI Companies

On the evening of March 29, I? ’s announcement showed that through the above transaction, Eli Lilly will obtain a global exclusive license to develop, produce, and commercialize a new oral therapeutic drug at the preclinical development stage for certain indications, with potential to reach a “best-in-class” level.

Industry speculation is that the molecule I? and Eli Lilly are developing together may be a GLP-1R target product. During the morning earnings call on March 30, materials presented by I? disclosed a pipeline line whose global rights have been authorized to an “undisclosed cooperation partner,” with the target being GLP-1R for treating metabolic diseases.

Eli Lilly has heavily bet on this area. In January, Eli Lilly’s collaboration with U.S. Nimbus Therapeutics was a development of a new oral treatment. Drug development is aimed at Eli Lilly’s core pipeline of obesity and other metabolic diseases. The upfront payment was $55 million (including recent milestone payments), with a potential total amount of up to $1.3 billion.

“Eli Lilly has money and feels free to spend.” An analyst from an investment institution believes this may be part of Eli Lilly’s strategy to build its weight-loss and glucose-lowering product portfolio: “Go your own way—so that others have no way to go.”

Buying out all globally promising products makes it harder for competitors to catch up.

Eli Lilly is currently the world’s most valuable pharmaceutical company, with a market capitalization of $837.7 billion in the U.S. stock market alone. Its dual-target weight-loss and glucose-lowering GLP-1 drug tirzepatide (not sure—keep original phrasing as English name) is the top-selling drug globally in 2025, with sales of $36.5 billion.

However, those sales are being closely chased by Novo Nordisk’s weight-loss and glucose-lowering GLP-1 drug semaglutide, which generated $36.1 billion in sales to rank second. In the coming years, competition in this track will become increasingly intense.

I? ’s GLP-1R target product has already completed lead compound optimization and entered the preparation phase for an Investigational New Drug (IND) application. In essence, Eli Lilly is buying a “R&D acceleration ticket.”

I? claims that using its Pharma.AI platform, it has compressed the drug R&D cycle for certain drugs from the traditional 3–6 years to within 18 months. Eli Lilly is buying exactly this “speed.” The two sides will also collaborate on multiple R&D projects around the targets selected by Eli Lilly.

Eli Lilly has strong interest in AI drug discovery. In just the past two years, it has initiated at least 10 external collaborations, covering therapeutic categories including muscle health, antibody design, and oncology drug development.

Eli Lilly’s best-known collaborator in AI is NVIDIA. In January 2026, the two sides jointly announced the launch of the first AI co-innovation laboratory, focused on using AI to tackle some of the most enduring challenges in the pharmaceutical industry; they will invest $1 billion in the future. In 2025, Eli Lilly collaborated with NVIDIA’s AI supercomputing system “AI factory” to accelerate drug discovery.

Ren Feng told Caijing, “Our main role is to provide high-quality services and delivery, with a relatively high success rate.”

No.2

China’s advantages in AI drug discovery

Compared with the earlier “PowerPoint story,” AI drug discovery entering 2026 is already ready for real-world execution.

In I? ’s pipeline, “now there are 28 PCCs (preclinical candidate drugs),” I? founder and CEO Alex Zhavoronkov introduced.

From another angle, business development (BD) is becoming an important source of revenue for China’s AI drug discovery companies as well, which also indicates their battle-tested capabilities.

In 2025, I? ’s revenue was $56.24 million (about RMB 389 million), of which both drug discovery and pipeline development accounted for more than 40%. From early 2026 to date, the company has already closed more than 10 external collaborations, with total value exceeding $4 billion.

Among the two “twin stars” of China’s AI drug discovery, another company, Genetron Holdings, took on in 2025 what is believed to be the largest single order in the history of AI drug discovery—about $5.99 billion in collaboration—with an upfront payment of $51 million. The counterparty is DoveTree Medicines. The collaboration includes multiple pipelines (oncology, immunology, neuroscience, and metabolism), an AI + robotics platform, and small molecules + antibodies.

China’s best-known AI success story is the January 2022 collaboration between Fosun Pharma and I? , widely seen as the starting point for China’s AI drug discovery to go international.

Under the agreement, the two sides jointly develop AI drugs targeting multiple targets worldwide. The deal value was $13 million upfront plus potential milestones. Although it was far smaller than Eli Lilly’s deal, it was the first time I? proved that its Pharma.AI platform could achieve large-scale commercialization licensing.

Alex Zhavoronkov, founder and CEO of I? , believes being in China is an important factor that allows one to frequently receive big deals from multinational pharmaceutical companies. Because even when AI is introduced to drug R&D, the test is not only computing power; it requires huge amounts of experimental data, which must be validated through cell experiments—not just algorithm models.

“Among the only two places in the world that can do efficient chemical synthesis, one is India and the other is China, and China’s quality is a bit higher.” Zhavoronkov said, “China has very excellent infrastructure and a great number of talents, which can help enable quick, large-scale validation. At the same time, the government’s support for these efforts is also very strong.”

Putting aside corporate self-promotion, multinationals’ confidence in China’s AI drug discovery is being built step by step.

Eli Lilly’s collaboration with Genetron Holdings began in 2019. By 2023, the two sides signed a $250 million AI + robotics small-molecule drug discovery collaboration. In 2025, they reached another multi-target collaboration totaling $345 million in the field of biologics drug R&D.

I? is similar. In 2022, Eli Lilly began subscribing to I? ’s software services. Three years later, in 2025, the two sides signed a project collaboration worth $100 million. Then, in March 2026, they completed a major collaboration worth $2.75 billion.

I? expects that in the first half and second half of 2026, more collaboration deals will be finalized and landed. Because some BD collaborations have already progressed to the stage of drafting cooperation frameworks.

“We’re working hard to turn these termsheets into contracts.” Ren Feng revealed that around the time the deal with Eli Lilly was announced and after, the company also received more partnership inquiries and invitations from overseas big pharma companies. He believes, “More deals will emerge than we imagine—especially with overseas multinational pharmaceutical companies.”

No.3

What’s next—will AI really start making drugs?

In late 2025, during I? ’s first public offering (IPO) phase, Eli Lilly subscribed to become a cornerstone investor, which was also the first time it participated in a Hong Kong listed biotech IPO as a company itself rather than through the Eli Lilly Asia fund. Therefore, after the antitrust review was triggered, rumors surfaced that Eli Lilly might acquire it.

Alex Zhavoronkov explicitly denied the rumors, saying, “Up to now, I? still wants to remain independent.” He added that if they were forced to talk about an acquisition, it could be a technology company—“for example, Microsoft or NVIDIA—the probability is higher than the probability of a pharmaceutical company acquiring us.”

Since it does not plan to be acquired for now, what I? needs to consider is how to achieve profitability.

“In the short term, our main revenue source is still BD revenue—upfront payments, milestone payments, revenue sharing, and so on. This will be the main business model at least for the next 3–5 years, or within 2–5 years.” Ren Feng told Caijing.

In the field of AI drug discovery, there has long been a dispute over business models. That is, whether AI is merely used as an aid for drug R&D, or whether it ultimately needs to truly produce a new drug with it.

The former is considered AI + CXO (pharma outsourcing), where you earn money by taking on outsourced R&D orders from pharmaceutical companies—this, to some extent, requires “low margins but high volume.” The latter is AI + Biotech, where you enjoy the larger profit space brought by pushing drugs to approval at the end.

Among China’s only two listed AI drug discovery companies, Genetron mainly follows the CXO route, while I? has consistently stated that it wants to make drugs itself.

This time, Ren Feng again reiterated I? ’s positioning of AI + Biotech. Ren Feng admitted, “We can never be an AI CRO (contract research organization). We want to provide services to customers, or sell projects to fund our own projects. We can’t rule out that, after 5 or 10 years in the future, we could further expand our diversified pipeline behind it.”

In I? ’s current pipeline, it has nearly 30 investigational new drugs, among which three independently developed products have advanced to the stage of clinical trials. The one with the fastest progress is an investigational product for treating idiopathic pulmonary fibrosis. According to the company’s management plan, the product will enter Phase 3 clinical trials in 2026.

Maybe the first drug developed with AI will be born in China.