OpenEvidence, the AI-powered medical search engine now used by more than a million licensed U.S. clinicians, is adding another layer to its oncology ambitions: a partnership with what it describes as a “nationally leading cancer center” that will bring expert-curated interpretation of cancer genomic alterations directly into the platform’s workflows, Fierce Healthcare first reported. OpenEvidence previewed the deal exclusively with the outlet, with plans to name the cancer center next week.
The move gives clinicians access to a precision oncology knowledge base — expert interpretation of genomic alterations paired with patient-specific context and peer-reviewed evidence — inside the same tool many already use at the point of care. Company founder Daniel Nadler, Ph.D., told Fierce Healthcare that OpenEvidence has actually lined up partnerships with several nationally leading cancer centers, with additional names to be announced “in the coming weeks.”
Oncology as the proving ground
The genomics partnership is the latest piece of an oncology stack OpenEvidence has been assembling for more than a year. The company holds a licensing agreement with the National Comprehensive Cancer Network that has digitized NCCN’s treatment algorithms directly into the platform, and it separately integrates ASCO guidelines, figures and flowcharts into a specialty oncology model. Earlier this year it struck a workflow-embedding deal with OneOncology, giving physicians across the network’s 565 sites access to evidence retrieval and clinical trial matching without leaving their existing applications, and it partnered with the Society of Surgical Oncology to bring surgical guidelines into the platform alongside a new research grant for AI-enabled clinical decision tools.
Taken together, the pattern is deliberate: rather than build a single generalist medical chatbot, OpenEvidence is stacking guideline bodies, provider networks and now genomics expertise specifically around cancer care, treating oncology as the first fully-realized specialty in a much larger architecture.
The “medical superintelligence” thesis
That architecture is what Nadler has been calling “medical superintelligence” — a system of AI agents, each functioning as a subspecialist, coordinated to act like a full multidisciplinary care team. He first laid out the vision publicly at the J.P. Morgan Healthcare Conference in January 2026, describing an approach modeled less on a single large language model and more on how a hospital, or an organization like NASA, distributes expertise across specialists who collaborate on hard problems.
Oncology is the first subspecialty OpenEvidence has built out this way, powered by an oncology sub-agent that has already digitized NCCN’s treatment algorithms and is now absorbing the incoming cancer center’s genomic knowledge base. Nadler told Fierce Healthcare that genetics, cardiology and neurology agents are next. “This dynamic ensemble of medically specialized agents will not only transcend the limitations of any individual human expert but will also result in a seismic-scale equalization of healthcare quality in America,” he said, arguing it could let physicians in rural or under-resourced counties carry what amounts to a round-the-clock, world-class tumor board in their pocket — for free.
Scale, and a new model family
The oncology partnership lands alongside a broader platform expansion. OpenEvidence now counts more than 1.12 million medical-licensed, NPI-verified U.S. clinicians as users — physicians, nurses, nurse practitioners and PAs — with August tracking toward more than 40 million verified clinician queries in a 30-day span, Nadler said. That reflects a steep growth curve for a company that reported roughly 430,000 verified physician users in mid-2025 and closed a $250 million Series D in January 2026 at a $12 billion valuation, backed by Thrive Capital, DST Global, Sequoia, Google Ventures, Nvidia and others.
The company also introduced a new family of models built around that superintelligence framing. Darwin, now in research preview, is OpenEvidence’s most advanced model and, the company says, the first AI system to post a perfect score on MedQA, alongside strong showings on MedXpertQA, HealthBench Professional and NOHARM — ahead of comparison models including Claude Fable 5 and Gemini 3.7. Darwin is available only to institutional partners such as the National Organization for Rare Disorders, research collaborators and accredited academic researchers for now, with its validated capabilities expected to flow down into the production models over time.
Those production models — Osler, Sackett and Snow, named for physicians credited with shaping modern medicine — began rolling out to all verified clinicians free of charge. Osler, the fastest at roughly five seconds per answer, replaces the model that currently powers OpenEvidence and remains the platform default; Sackett is a deeper evidence-weighing model; and Snow, the slowest and most thorough, runs a fuller review of the medical literature before returning a report. “Every model in the family is held to the same standard of clinical accuracy,” the company said. “What varies is time: how long a model thinks, and how deep it searches.”
Why it matters for oncology
For cancer programs and practices, the near-term signal is less about any single feature than about direction: a company already embedded in a large share of U.S. clinical workflows is making oncology, and specifically precision oncology genomics, its flagship specialty rather than a bolt-on feature. Whether the promised “seismic-scale equalization” of care materializes will depend on how well genomic interpretation, guideline bodies and point-of-care speed hold up outside a demo — but the direction of investment is clear, and more cancer-center names are expected within weeks.