Recursion OS: Inside the Lab-in-the-Loop Engine Rewriting Biology, Chemistry and Clinical Prediction

Overview: How Recursion OS’s AI-Driven End-to-End Drug Discovery and Development Platform Transforms Pharma R&D

Recursion OS is an end-to-end AI platform for drug discovery and development that combines automated experimentation, multimodal biological and chemical data, machine learning, and computational chemistry to support target discovery, hit identification, molecule design, and translational research. It addresses the high cost, long timelines, and attrition associated with sequential, hypothesis-led drug development by creating an integrated feedback loop between laboratory experiments, data generation, and in silico analysis.

The platform aggregates high-content cellular imaging, genomics, transcriptomics, proteomics, ADME and toxicology data, chemical libraries, and de-identified patient data into a connected research environment. Its computational tools model relationships among biological perturbations and compounds, enabling researchers to prioritise targets, identify potential on- and off-target effects, expand hits through structure- and similarity-based methods, and guide candidate design for further experimental validation. Recursion describes a “lab-in-the-loop” workflow in which physical experiments and machine-learning models continuously inform one another.

For R&D teams, this model can consolidate otherwise fragmented experimental and computational workflows, enabling broader hypothesis generation and more structured prioritisation of programmes before resource-intensive downstream studies. Recursion states that its automated infrastructure can orchestrate up to 2.2 million experiments per week and that its platform can identify hit candidates in weeks rather than years; realised timelines, cost savings, and programme outcomes will depend on the biology, assay design, chemistry, validation strategy, and clinical evidence for each programme.

Last checked on September 30, 2026: Recursion OS remains active as an end-to-end AI drug-discovery platform, with Genentech exercising its first validated neuroscience-target option under the companies’ collaboration in August 2026.

What is Recursion OS?

Recursion OS is an AI-native, end-to-end drug-discovery and development platform that combines automated wet-lab experimentation, multimodal biological and chemical data, machine learning, and computational chemistry to support target discovery, hit identification, molecule design, and translational research. It is used by Recursion’s internal R&D teams and pharmaceutical partners working across oncology, rare disease, neuroscience, immunology, and other therapeutic areas. Its distinguishing approach is a lab-in-the-loop system that links large-scale automated experiments with proprietary data assets and AI models; Recursion also publishes selected datasets and models for non-commercial research use.

Intended Use and Limitations

• Recursion is a Nasdaq-listed clinical-stage TechBio company operating Recursion OS, an AI-enabled drug-discovery and development platform.

• Recursion completed its acquisition of Exscientia in November 2024, combining Recursion’s automated biological experimentation and data infrastructure with Exscientia’s AI-enabled chemistry and drug-design capabilities.

• Recursion has a strategic collaboration with Roche and Genentech that uses Recursion OS and biological maps to identify targets in neuroscience and oncology; the partnership may initiate up to 40 programmes over a decade or longer.

• Genentech advanced the first neuroscience target from the Recursion collaboration into early small-molecule discovery in 2026.

• Recursion’s Sanofi collaboration covers AI-driven small-molecule discovery in immunology and oncology; Recursion reported five achieved milestones and $134 million in upfront and progress-based payments by February 2026.

• Recursion has an expanded precision-oncology collaboration with Bayer covering up to seven programmes, with potential development and commercial milestones plus royalties if programmes progress successfully.

• Recursion’s platform combines automated wet-lab experiments with machine learning, biological and chemical datasets, and computational chemistry in an iterative lab-in-the-loop model.

• Recursion reported preliminary Phase 2 results for REC-4881 in familial adenomatous polyposis and described the programme as a clinical validation of an insight generated through its full-stack AI operating system; this is company-reported programme evidence.