- New “AI GYM for Science” dramatically boosts the biological and chemical intelligence of any causal or frontier LLM
- Up to 10x performance gains on key drug discovery benchmarks compared to LLMs that miss the mark on ~75–95% of tasks
- Advances Insilico’s vision of Pharmaceutical Superintelligence (PSI)
CAMBRIDGE, Mass., Jan. 22, 2026 /PRNewswire/ — Insilico Medicine (“Insilico”, HKEX: 3696), a leading global AI-driven biotech company with a clinical-stage pipeline, today announced the launch of Science MMAI Gym, a domain-specific training environment designed to transform any causal or frontier Large Language Model (LLM) into a high-performance engine for real-world drug discovery and development tasks.
Building on more than a decade of AI research and its own internal pipeline of 27 preclinical candidates, 10+ molecules with IND clearance, and multiple Phase I and Phase IIa clinical trials completed or ongoing, Insilico is now opening its AI training infrastructure to external partners. Science MMAI Gym adapts general-purpose LLMs (such as GPT, Claude, Gemini, Grok, Llama, and Mistral) to reason in medicinal chemistry, biology, and clinical development with the precision required in modern pharma R&D.
Alex Zhavoronkov, PhD, founder, CEO and CBO of Insilico Medicine invites pharma and biotech companies, AI labs, and cloud providers to “bring your AI model to Science MMAI Gym” and explore CSI, BSI, and PSI memberships tailored to their R&D pipelines.
Addressing the LLM Performance Gap in Drug Discovery
Despite their general intelligence, flagship LLMs often fail or underperform on mission-critical drug discovery tasks, such as predicting complex pharmacokinetic (DMPK) and toxicity endpoints (e.g., hERG, DILI risk, LD50). Insilico’s benchmarks show that without specialized training, general models often produce vague or chemically implausible reasoning, even with advanced prompting.
Science MMAI Gym directly addresses this by teaching LLMs domain-specific scientific reasoning—the language, formats, and conceptual chains used by chemists and biologists rather than treating drug discovery as a simple NLP benchmark.
The Gym’s curriculum focuses on:
- Medicinal and organic chemistry: Multi-step optimization chains, reaction reasoning, retrosynthesis, and 3D structure-property relationships.
- Biology and target discovery: Omics-aware reasoning over gene expression, pathways, disease mechanisms, and multi-objective target scoring.
- Clinical development: Interpretation of trial designs, endpoints, and prediction of Phase 2 trial success or failure using proprietary benchmarks like ClinBench.
The Training Architecture: CSI and BSI
Science MMAI Gym is a core component of Insilico’s roadmap toward Pharmaceutical Superintelligence (PSI), with dedicated tracks for Chemical Superintelligence (CSI) and Biology/Clinical Superintelligence (BSI).
Models “train” in the Gym over a period of weeks to months using:
- High-quality, domain-specific reasoning datasets: Leveraging millions of internal data points, including over 4 million medicinal chemistry optimization chains, 100 million organic synthesis descriptions, and hundreds of thousands of molecular dynamics trajectories.
- Multi-task Fine-Tuning and Reinforcement Learning: Models undergo Multi-task Supervised Fine-Tuning (SFT) and Reinforcement Fine-Tuning (RFT) using reward models to hone reasoning skills and align predictions with experimental outcomes.
- Robust Benchmarking: Each cycle is evaluated against public and in-house out-of-distribution (OOD) benchmarks, including TDC, TargetBench, and ClinBench, ensuring the model’s performance is robust for real-world application.
Case Studies: From Generalist to Specialist
Internal benchmarks demonstrate the transformative effect of the Gym:
- Chemistry (CSI): A tested open-source causal LLM, which previously failed on 70% of medchem tasks, emerged from the Gym as a “single-model-does-it-all” chemistry engine. It achieved SOTA or near-SOTA performance on multiple ADMET tasks and delivered SOTA Success Rate on five optimization tasks in the MuMO-Instruct benchmark, matching or exceeding strong category-specific generalist models.
- Biology/Clinical (BSI): Models trained at the Gym demonstrated substantial gains on proprietary benchmarks. On ClinBench, a model’s F1 score for predicting Phase 2 trial outcomes rose significantly, outperforming a broad set of frontier LLMs. Similarly, on TargetBench, BSI-tuned models achieved the top composite ranking for novel target identification across multiple diseases, demonstrating high biological plausibility and translational readiness.
Business Model: “Membership” in the AI GYM for Science
Science MMAI Gym is offered as a flexible, membership-style program, ranging from intensive two-week sprints to longer PSI-oriented engagements. Partners provide their base model and receive a CSI/BSI/PSI-enhanced version with up to 10x performance improvement compared to baseline, along with detailed benchmark reports and optional wet-lab validation through Insilico’s automated assay platforms.
About Insilico Medicine
Insilico Medicine is a pioneering global biotechnology company dedicated to integrating artificial intelligence and automation technologies to accelerate drug discovery, drive innovation in the life sciences, and extend health longevity to people on the planet. The company was listed on the Main Board of the Hong Kong Stock Exchange on December 30, 2025, under the stock code 03696.HK.
By integrating AI and automation technologies and deep in-house drug discovery capabilities, Insilico is delivering innovative drug solutions for unmet needs including fibrosis, oncology, immunology, pain, and obesity and metabolic disorders. Additionally, Insilico extends the reach of Pharma.AI across diverse industries, such as advanced materials, agriculture, nutritional products and veterinary medicine. For more information, please visit www.insilico.com
