OpenAI expands GPT-Rosalind access for academic research

OpenAI has launched the Rosalind Academic Plan, expanding access to its specialized biological research AI model, GPT-Rosalind, to eligible academic institutions. The program is designed to support principal investigators and laboratory leaders in fields such as biology, drug discovery, and translational medicine.

Under the plan, eligible US university labs may receive up to 30 Business Standard seats for a period of 12 months. To address privacy concerns common in scientific research, OpenAI stated that research data from these accounts will not be used to train its models by default. Access is not automatic and requires organizational approval based on the legitimacy of the scientific research and governance standards.

GPT-Rosalind is engineered to reason across biological entities, synthesize scientific literature, plan experiments, and connect with scientific databases. The model has demonstrated performance improvements on benchmarks such as Genebench and Medchem Bench. The broader ecosystem includes the Rosalind Workbench, a tool intended to streamline research workflows by connecting scientific tools with data analysis procedures.