Insilico Medicine

Learning Objectives

• Understand Insilico Medicine's full-stack AI drug discovery platform
• Explain why rentosertib's Phase III trial is the key test of AI-designed drugs
• Evaluate when Insilico fits a pharma partnership vs IsoDDE or BenevolentAI

What Is Insilico Medicine?

Insilico Medicine is one of the most clinically advanced AI drug discovery companies. It operates the Pharma.AI suite, which covers target identification, generative chemistry, and clinical-trial prediction, and it uses that software on its own drug pipeline as well as licensing it to partners.

The strategic significance is its lead drug, rentosertib, for idiopathic pulmonary fibrosis. Insilico's AI found the target and designed the molecule, and on September 10, 2026 the first patient was dosed in a Phase III trial. Insilico calls it the world's first Phase III trial of a drug discovered and developed with generative AI. Most newer AI drug discovery companies are still preclinical or in early human trials, so the outcome will be one of the clearest tests yet of whether the approach produces better medicines.

Insilico is also now a public company. It listed on the Hong Kong Stock Exchange Main Board on December 30, 2025, under stock code 3696, raising about 2.277 billion Hong Kong dollars, the largest biotech listing in Hong Kong that year.

Status & Pipeline

Insilico's clinical progress and its pharma deals are the company's two commercial proof points.

Insilico's revenue model combines partnership payments (upfront fees, milestones and royalties) with the potential value of its own drugs if they reach the market. Pricing is not published; every engagement is negotiated.

Core Capabilities

Pharma.AI Platform Suite

The end-to-end drug discovery AI platform:

• PandaOmics — target identification using multi-omics data
• Chemistry42 — generative chemistry for novel molecule design
• inClinico — prediction of clinical-trial success

These three modules cover the drug discovery journey from disease biology to clinical trial planning. Newer modules extend the suite to antibody and protein design and to research assistance.

Rentosertib: From AI Target to Phase III

The headline credibility differentiator. Insilico's AI identified an enzyme called TNIK as a new target in idiopathic pulmonary fibrosis, a progressive scarring of the lungs, and Chemistry42 designed rentosertib to block it. The drug has now passed each stage a conventional drug must:

• IND-enabling studies and Phase I safety
• A Phase II trial, with results published in Nature Medicine in 2025 showing improved lung function against placebo
• GENESIS-IPF-3, a 52-week Phase III trial of about 320 patients at 47 centers in China, with the first patient dosed on September 10, 2026

Phase III is where most drug candidates fail, so rentosertib is a test of the approach rather than a proof of it. Behind it, Insilico reported more than 30 programs at its listing, ten of them cleared for human testing.

Generative Chemistry

Chemistry42 uses generative AI to design entirely new molecules with desired properties — rather than screening existing libraries. Generates candidates predicted to bind specific targets with optimized ADMET (Absorption, Distribution, Metabolism, Excretion, Toxicity) properties.

Target Identification (PandaOmics)

The disease-biology layer. PandaOmics integrates multi-omics data (genomics, transcriptomics, proteomics, etc.) and biomedical literature to identify novel disease targets — proteins or pathways that haven't been previously considered druggable.

Clinical Trial Prediction (inClinico)

The trial-planning layer. inClinico uses AI to predict clinical trial outcomes, estimating how likely a trial is to succeed from the data available before it starts. Helps sponsors spot weak programs before committing to an expensive study.

Pharma Partnership Track Record

Insilico says 10 of the top 20 global pharmaceutical companies work with its platform, and its disclosed deals have grown in size:

• Sanofi (2022) — a discovery collaboration worth up to 1.2 billion dollars
• Eli Lilly (March 2026) — 115 million dollars upfront and about 2.75 billion dollars in total, plus royalties, for an exclusive worldwide license to oral drug candidates still in preclinical development
• Takeda (July 2026) — about 60 million dollars near term and about 600 million dollars in total, plus royalties, applying Pharma.AI across six Takeda therapeutic areas

Eli Lilly was also a cornerstone investor in the Hong Kong listing.

Strengths

• Phase III lead drug: Rentosertib is, by Insilico's account, the first generative-AI-discovered drug to reach Phase III
• Full-stack Pharma.AI platform: Target ID + generative chemistry + trial prediction
• Large pharma partnerships: Eli Lilly, Takeda and Sanofi deals, with 10 of the top 20 pharma companies on the platform
• Generative chemistry depth: Produces entirely new molecules
• Multi-omics target identification: PandaOmics integrates diverse data sources
• Public-market funding: The 2025 Hong Kong listing funds the internal pipeline

Limitations & Considerations

• Public-company scrutiny: Listed in Hong Kong; reported results and trial news move the stock
• Phase III is not an approved drug: Most Phase III candidates still fail, and rentosertib's trial runs only in China so far
• Pharma partnership cycles: Multi-year deals with lumpy, milestone-driven revenue
• Custom-quote pricing: Not transparent
• Competition intensifying: IsoDDE, Manas AI, others advancing rapidly
• Less recent architectural novelty: Insilico predates AlphaFold 3-derived approaches

Best Use Cases

When to choose alternatives:

• Most-novel architectural approaches → Isomorphic Labs IsoDDE for AlphaFold 4-class structure prediction
• Cellular reprogramming research → Altos Labs for longevity-specific work
• Knowledge-graph drug discovery → BenevolentAI
• Physics-based drug discovery → Schrödinger combines physics + ML
• Specialized therapeutic-area AI biotech → category-specific competitors

Key Takeaways

• Insilico Medicine is one of the most clinically advanced AI drug discovery companies: its lead drug, rentosertib, entered a Phase III trial in September 2026, which Insilico calls the first for a generative-AI-discovered drug
• Pharma.AI suite covers the drug discovery journey: PandaOmics (target identification), Chemistry42 (generative chemistry), inClinico (clinical trial prediction)
• Pharma partnership model with upfront, milestone and royalty economics; recent deals with Eli Lilly (up to about 2.75 billion dollars) and Takeda (about 600 million dollars), after Sanofi in 2022
• Listed on the Hong Kong Stock Exchange Main Board in December 2025, the largest biotech listing in Hong Kong that year
• Best fit for pharma drug discovery partnerships and as the leading AI-to-clinic case study; for cutting-edge architectures use IsoDDE; for cellular reprogramming use Altos Labs; for knowledge-graph approaches use BenevolentAI