OpenAI is handing US university biology labs a free pass to its most specialized science model. On October 15, 2026, the company unveiled the Rosalind Academic Plan, which offers sponsored access to GPT-Rosalind for eligible academic researchers.
What the Rosalind Academic Plan actually offers
The program targets principal investigators and lab leaders at US universities. Eligible labs can receive up to 30 free Business Standard seats for 12 months.
Those seats unlock GPT-Rosalind along with other advanced OpenAI models.
OpenAI also says research data from these accounts will not be used to train its models by default. Academic labs often work with unpublished findings, sensitive datasets, and ideas they would rather not see resurfacing in someone else’s chatbot answer. A default opt-out removes one of the bigger objections researchers raise about using commercial AI tools.
Access is not automatic. Organizational approval is required, and OpenAI says the focus is on legitimate scientific research.
What GPT-Rosalind brings to the bench
GPT-Rosalind is designed to reason across biological entities, synthesize evidence from scientific literature, plan experiments, and connect with established scientific databases.
Compared with existing baselines, GPT-Rosalind posted performance gains of:
53.7% on Genebench, 18.0% on Medchem Bench, and 19.6% on Labworkbench.
The broader Rosalind brand also includes the Rosalind Workbench, a tool meant to streamline research workflows.
A model called gpt-rosalind-discovery was listed at $5 per million input tokens as of October 10, 2026.
From enterprise preview to campus rollout
GPT-Rosalind first launched on April 16, 2026, as a research preview. Early access was limited to vetted enterprise and research organizations.
OpenAI has highlighted early collaborations with Amgen and Moderna, pointing to potential applications in areas including biodefense and public health. OpenAI has emphasized responsible deployment with governance oversight, which helps explain why access still runs through organizational approval rather than a simple sign-up button.
Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.