OpenAI has introduced the Rosalind Academic Plan, giving eligible university labs up to 30 Business Standard seats free for up to 12 months, with approval required to use its life sciences model, GPT‑Rosalind. The plan is currently open to labs at U.S. universities and includes Rosalind Workbench, which coordinates specialized tools for research tasks such as genomics and medicinal chemistry. Access also covers OpenAI’s commercially available frontier models, within the plan’s usage limits.
Principal investigators and lab leads whose universities lack an OpenAI Enterprise agreement can submit an application for their lab through the trusted access process. Researchers at institutions that already have an agreement must use their university’s Enterprise access process instead. Approved leads can activate access for up to 30 named lab members, but the plan excludes API access.
Read more: OpenAI Offers US Government 5 Percent Stake to Defuse Political Pressure Ahead of IPO
OpenAI says research data submitted under the Rosalind Academic Plan is excluded from model training by default. Verified principal investigators and lab leads control member permissions and take responsibility for the lab’s authorized use. Before access begins, OpenAI checks the institution and lead’s identity, examines the lead’s appointment and oversight authority, and evaluates the lab’s research and intended applications.
OpenAI describes GPT‑Rosalind as a model built to reason through difficult biological problems, with approved researchers able to explore sensitive questions under research safeguards. Potential applications include developing candidate medicines and therapeutic proteins, studying resistance to antimicrobial treatments, and investigating ways to detect disease with less invasive tests. OpenAI says academic labs are already exploring the model for protein design and research into neurodegeneration.
The approval process also assigns responsibilities to the lab lead and an institutional contact responsible for safety. OpenAI says monitoring and enforcement continue after activation, and researchers may be asked to explain their work or confirm that it is authorized. For registered research involving heightened sensitivity, the company says its checks take the approved scientific context into account.
The API exclusion distinguishes this offer from GPT-6 Astra Ultrafast, which became available through the OpenAI API earlier this month and also reached eligible ChatGPT Work and Codex users. OpenAI separately began bringing GPT-6 with Intelligent UI to ChatGPT on October 7, with interactive responses for everyday conversations.
