AI drug company liked by Jensen Huang, Insilico Medicine, IPOs on Hong Kong Stock Exchange today.
```
“In the future, all biology will start with ‘computer simulation’ (in silico) and, to a great extent, end with ‘computer simulation’.” (Almost everything will largely start in silico, largely end in silico)
When Jensen Huang, Founder and CEO of NVIDIA, repeatedly stated this assertion under the spotlight at the J.P. Morgan Healthcare Conference and GTC Conference, biology was transforming from an experimental science relying on trial and error into a predictable, programmable data science.
Insilico (Insilico Medicine) is precisely the “first principle” practitioner repeatedly mentioned by Jensen Huang.named
On December 30, 2025, Insilico (3696.HK) was officially listed on the Main Board of the Hong Kong Stock Exchange, opening at HKD 35 per share, up 45% from the issue price, with a market capitalization reaching HKD 19.5 billion.
This is the largest IPO event in the biopharma sector of Hong Kong stocks in 2025, raising a total of HKD 2.277 billion; it is also a test of the “AI+Biotech” business model in the capital market.
Unlike the numerous unprofitable biotech companies listing under Chapter 18A, Insilico is the first AI biopharma company to list under Article 8.05 of the Main Board listing rules—this means the company not only holds expectations for its future pipeline, but has also passed rigorous profit or revenue tests, thus demonstrating actual commercial viability.
The listing of Insilico marks the watershed moment as the AI pharmaceutical sector moves from “proof-of-concept” to “industrial-scale output”.
01 Capital Market “Voting with Their Feet”: Luxury Lineup Led by Eli Lilly + Tencent
According to the IPO results, the Hong Kong public offering portion saw about 1,427.37 times oversubscription, locking in more than HKD 328.349 billion; the international placement was also oversubscribed by 26.27 times. Both figures have set records for non-Chapter 18A healthcare IPOs in Hong Kong this year.
In Hong Kong IPOs, the cornerstone investor list reflects institutional judgment of an issuer’s fundamentals. This time, Insilico attracted 15 global cornerstone investors, with a total subscription of about US$115 million.
The most striking names are Eli Lilly and Tencent.
This is the first time Eli Lilly has bet on the AI pharmaceutical track as a cornerstone investor. This sends a strong signal: MNCs (multinational pharma companies) not only recognize Insilico’s technology platform, but are paving the way for future pipeline collaboration.
Similarly, this is the first time Tencent has participated as a cornerstone investor in a Biotech IPO. This represents tech giants’ endorsement of the “AI+Science” cross-disciplinary integration trend. AI pharmaceuticals require huge computing power and cloud infrastructure, and Tencent may provide not only capital, but also deeply integrated computing infrastructure.
Meanwhile, Oaktree Capital returns to the Hong Kong Biotech market for the first time this year. Known for its expertise in distressed assets and value discovery, Oaktree’s selection of Insilico as its “first shot” may signal the company’s risk-reward profile is now sufficiently attractive.
Other cornerstone investors include Temasek, Schroders, UBS, E-Fund, and Taikang Life.
This “feverish” subscription enthusiasm confirms its AI premium. Since ChatGPT ignited the generative AI boom, AI concept stocks have enjoyed huge premiums and Insilico is reflecting this valuation expectation.
On its first day of trading, Insilico opened 45% above its issue price, showing that even during a biopharma capital winter, the market remains willing to give “hard tech” a high valuation tolerance.
02 Business Model Restructuring: The Flywheel Effect of “Dual Engines”
The market’s enthusiasm for Insilico centers on its unique “dual engine” business model: artificial intelligence + innovative drug discovery.
Insilico licenses its proprietary generative AI platform Pharma.AI to pharmaceutical companies, charging subscription fees. This brings not only predictable recurring revenue (ARR), but also extremely high client stickiness.
Once pharma companies get used to using Chemistry42 for molecule generation, their switching costs become extremely high. This creates a natural client pool for subsequent pipeline collaborations. Meanwhile, widespread software deployment allows Insilico to collect large amounts of external user feedback data, feeding back to its algorithms and creating a “data-algorithm-product” closed loop.
Since 2020, the Pharma.AI platform has been commercialized as a modular software product, with a global partner network. As of the last practicable date, it has signed software licensing deals with 13 of the top 20 pharmaceutical companies worldwide.
Innovative drug discovery is the real engine of explosive growth.
This is the typical Biotech model—but with greater efficiency. Insilico uses its own platform to develop new drugs, generating income via licensing (License-out) or self-development to clinical stages.
This segment currently contributes more than 90% of its revenue.
The core logic: By leveraging AI’s high success rate, it produces batches of preclinical candidate drugs (PCCs), monetizing them at high-value points. This SaaS + Biotech hybrid solves the pain point of zero revenue and high risk at early-stage Biotech IPOs, while retaining massive valuation flexibility.
This is the key valuation premium that sets Insilico apart from traditional CXOs and pure Biotechs.
03 The Power of AI Pharmaceuticals: 12-18 Months vs. 4.5 Years
The AI drug discovery track doesn’t lack for stories; what’s missing is validation by clinical data. Insilico’s biggest moat is: proving AI efficacy with clinical data.
According to the Frost & Sullivan report, traditional drug discovery takes around 4.5 years from target identification to PCC nomination. With Pharma.AI, Insilico slashes this process to 12-18 months, with each project requiring synthesis and testing of only 60-200 molecules.
With the same budget in terms of money and time, Insilico can try more targets and accumulate more trial opportunities. For drug R&D, which is a “high risk, high reward” gamble, AI systematically changes the odds by increasing the number of bets and improving single-shot probability.
ISM001-055 (Rentosertib) is the best annotation of this logic.
It’s the world’s first clinical-phase II candidate drug whose target and molecule design were discovered by AI, targeting idiopathic pulmonary fibrosis (IPF). PandaOmics identified TNIK as a potential target, and Chemistry42 generated a new molecular structure, realizing a complete closed loop.
Phase IIa topline data released in October 2024 showed positive efficacy signals in patients, with excellent dose-dependency. The trend of FVC (forced vital capacity) improvement validated the accuracy of AI predictions.
The success of ISM001-055 completes the pharmaceutical industry’s “Turing test”—it proves that AI not only can generate molecular structures, but the resulting drugs are truly safe and effective in human bodies.
In addition, the company has built up a solid pipeline.
ISM3091 (USP1 inhibitor) has been licensed to Exelixis for a total transaction worth US$955 million; ISM5043 (KAT6 inhibitor) has been licensed to Stemline (a subsidiary of Menarini); ISM5411 (PHD1/2 inhibitor), developed in-house for IBD, is also now in Phase I clinical trials. These pipelines not only prove the replicability of the AI platform, but also provide the company with ongoing “blood-making” capability.
04 Conclusion
The past decade saw the miracle of Moore’s Law in the chip industry. Perhaps the next decade will see AI overturn “Eroom’s Law” (the anti-Moore’s Law, describing how drug R&D costs double exponentially over time).
Under Jensen Huang’s “In Silico” prophecy, Insilico has taken a critical step. Whether this step will bring truly accessible, affordable, and breakthrough therapies to patients worldwide—time will tell.
Risk Reminder and DisclaimerThe market involves risk; investments should be made cautiously. This article does not constitute individual investment advice and does not take into account the special investment objectives, financial circumstances, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article suit their particular circumstances. Investments made accordingly are at your own risk. ```