OpenAI has introduced GPT-Rosalind, a new AI model designed for life sciences research, with a focus on biology, drug discovery, and translational medicine. The company says the system is intended to help researchers work through scientific literature, data, tools, and experimental workflows more efficiently.
The model arrives at a time when drug development remains a lengthy process. OpenAI said it can take roughly 10 to 15 years for a new medicine to move from target discovery to regulatory approval in the United States. The company argues that improvements made early in discovery can shape later results, including target selection, hypothesis quality, and experiment design. It says GPT-Rosalind is aimed at helping scientists move faster through those early stages.
OpenAI described GPT-Rosalind as a frontier reasoning model tuned for scientific tasks. The company said it is optimized for work across chemistry, protein engineering, and genomics, and that it performs particularly well in multi-step workflows that involve literature review, sequence interpretation, experimental planning, and data analysis.
According to OpenAI, the model is the first release in a broader life sciences model series. The company said future versions will continue to improve biochemical reasoning and support long-horizon research tasks that depend heavily on external tools and databases.
GPT-Rosalind is available as a research preview in ChatGPT, Codex, and the API for qualified customers through OpenAI’s trusted access program. The company is starting with U.S. enterprise users under a controlled deployment model that includes eligibility checks, access management, and governance requirements.
OpenAI also launched a free Life Sciences research plugin for Codex. The plugin is designed to connect models to more than 50 scientific tools and data sources, covering areas such as human genetics, functional genomics, protein structure, biochemistry, clinical evidence, and study discovery. OpenAI said the plugin can be used with its mainline models more broadly, while eligible enterprise customers can combine it with GPT-Rosalind for deeper biological reasoning.
To assess the model, OpenAI said it tested GPT-Rosalind on tasks tied to scientific discovery, including chemical reaction mechanisms, protein structure, mutation effects, DNA sequence interpretation, experimental output analysis, and follow-up experiment design. The company said these evaluations suggest progress in end-to-end research workflows.
OpenAI also cited benchmark results. On BixBench, a bioinformatics and data analysis benchmark, it said GPT-Rosalind posted leading performance among models with published scores. On LABBench2, which measures research tasks such as literature retrieval, database access, sequence manipulation, and protocol design, OpenAI said the model outperformed GPT-5.4 on 6 of 11 tasks.
The company said it also worked with Dyno Therapeutics to test the model on RNA sequence-to-function prediction and generation using unpublished sequences. In Codex, OpenAI said best-of-ten submissions from the model ranked above the 95th percentile of human experts on the prediction task and around the 84th percentile on the generation task.
OpenAI said it is already working with companies and research groups including Amgen, Moderna, the Allen Institute, and Thermo Fisher Scientific. Amgen’s senior vice president of AI and data, Sean Bruich, said the collaboration could help accelerate how medicines reach patients, while noting that life sciences work requires precision and that the data involved are highly complex.
OpenAI said it is limiting access to qualified organizations conducting legitimate scientific research with public benefit. The company emphasized enterprise-grade security controls, oversight, and misuse prevention as part of the launch.
Organizations can request access through OpenAI’s qualification and safety review process. During the research preview, the company said the model will not consume existing credits or tokens, subject to abuse guardrails.
The model is named after Rosalind Franklin, whose research helped reveal DNA’s structure and shaped modern molecular biology. OpenAI said the launch marks the beginning of a longer effort to build AI systems that can support scientific discovery in health and biology.