Owkin and NVIDIA Forge Alliance for Biological Superintelligence

SAN FRANCISCO & PARIS – January 12, 2026

AI biology company Owkin has announced a landmark collaboration with NVIDIA, uniting to enhance its biological large reasoning model, OwkinZero, in a bold move toward creating what it terms a Biological Artificial Super Intelligence (BASI). The partnership aims to leverage NVIDIA's formidable AI computing ecosystem to accelerate breakthroughs in drug discovery and development, potentially reshaping the multi-billion dollar pharmaceutical research landscape.

This collaboration marks Owkin's first with the computing giant and represents a significant step in constructing the sophisticated agentic infrastructure required to understand and manipulate complex biological systems. By integrating NVIDIA's technology, Owkin seeks to dramatically improve the performance, scalability, and robustness of its AI, which is designed to reason about biology in ways that were previously unattainable.

The Quest for Biological Superintelligence

At the heart of the announcement is the ambitious goal of building a Biological Artificial Super Intelligence. Owkin conceptualizes BASI not merely as a more powerful algorithm, but as an AI system capable of reasoning across the full spectrum of biological complexity. The company's strategy involves creating an 'AI scientist' that can automate research, understand causal biology, and ultimately discover and develop new diagnostics and therapies with minimal human intervention.

This vision moves beyond current AI applications in medicine, which often focus on discrete tasks like image analysis or data classification. Owkin aims to build an agent that can form and test hypotheses, integrate disparate data types, and generate novel biological insights.

"Building the first Biological Artificial Super Intelligence requires not just powerful models, but the right agentic infrastructure and toolkits to scale them," said Thomas Clozel, M.D., co-founder and CEO of Owkin, in the company's press release. "Our collaboration with NVIDIA represents a crucial step in developing the intelligent infrastructure for biology. By combining NVIDIA's world-class AI ecosystem with our domain-specific biological reasoning capabilities and multimodal patient data, we'll be able to create AI agents that can truly understand and reason about the complexity of biology at a scale that was previously impossible."

This work will directly power Owkin K, the company's AI copilot for researchers, and its internal AI Scientist, automating the path to new cures.

Powering Discovery with Advanced AI

The technical core of the collaboration involves integrating OwkinZero with NVIDIA's most advanced tools. Owkin will leverage the NVIDIA Nemotron family of open models, a suite of powerful AI architectures designed for building specialized and efficient AI agents. The Nemotron family includes models of varying sizes, from the 30-billion-parameter Nemotron 3 Nano, designed for high-throughput tasks, to the 500-billion-parameter Nemotron 3 Ultra, intended for applications demanding the highest reasoning accuracy. This tiered approach allows for tailored application, from rapid analysis to deep, complex problem-solving.

Alongside Nemotron, Owkin will utilize the NVIDIA NeMo framework, an open-source platform for building and training deep learning models, particularly within the biopharma sector. A key component is NVIDIA NeMo RL, a library for reinforcement learning that will help Owkin customize and scale OwkinZero's performance. This technology enables the model to learn and improve from its interactions, refining its biological reasoning capabilities over time.

Owkin's existing AI agents have already shown impressive efficiency. Its Pathology Explorer agent, for instance, achieved 23.7% better classification accuracy on key benchmarks while using five times fewer parameters than comparable models, reducing computational timelines from weeks to hours. The infusion of NVIDIA's technology is expected to amplify these gains significantly.

The Data Engine: From Patient to Prediction

A key differentiator for Owkin, and the fuel for its ambitious AI, is its exclusive access to a vast and unique repository of proprietary multimodal patient data. Gathered over a decade from a global network of over 800 hospitals and 104 healthcare centers, this dataset provides a rich, real-world foundation for training AI models that is difficult for competitors to replicate.

This 'multimodal' data includes a wide array of information, such as gene mutational status, tissue histology, single-cell gene expression, and clinical patient outcomes. By training its AI on this complex, integrated data from real human patients, Owkin aims to generate insights that more accurately reflect human disease compared to models trained on siloed or purely laboratory-generated data.

Recognizing the sensitivity of this information, Owkin emphasizes a robust ethical and privacy framework. The company employs a 'privacy by design' approach, adhering to strict data protection regulations like GDPR. All patient data is pseudonymized or de-identified before Owkin's researchers access it, with the link between a patient's identity and their data code held securely by the originating hospital. This ensures that individual privacy is protected while still allowing for large-scale scientific analysis. This commitment to ethical data handling is crucial for building trust with both medical partners and the public as AI plays an ever-larger role in healthcare.

Reshaping a Competitive Landscape

Owkin's partnership with NVIDIA arrives as the AI in drug discovery market is experiencing explosive growth. Market forecasts project the sector to grow from just a few billion dollars today to over $25 billion by 2030, with a compound annual growth rate (CAGR) exceeding 40%. Companies like BenevolentAI, Insilico Medicine, and Exscientia are all vying for leadership in this space, each using proprietary AI platforms to accelerate target discovery and molecule generation.

In this crowded field, Owkin's strategy is to leverage its unparalleled clinical data access as a primary competitive advantage. While many rivals focus on AI for chemistry or preclinical biology, Owkin's 'patient data-first' approach allows it to prioritize drug targets with a higher probability of clinical success. This has already helped the company forge partnerships with eight of the world's top ten largest pharmaceutical companies.

By combining this unique data asset with NVIDIA's state-of-the-art computing power, Owkin is positioning itself not just as a participant but as a potential leader in the next generation of biopharmaceutical research. The collaboration aims to create a feedback loop where better AI leads to deeper biological insights, which in turn informs the development of more effective therapies. This strategic alignment of biological data, AI reasoning, and accelerated computing sets a new benchmark in the race to automate scientific breakthroughs and deliver novel therapies to patients worldwide.