OpenAI · 2026-04-16 · major
GPT-Rosalind — OpenAI's first domain-specific frontier reasoning model, for life sciences
OpenAI ships GPT-Rosalind, a frontier reasoning model purpose-built for life sciences — evidence synthesis, hypothesis generation, experimental planning. Research preview via ChatGPT, Codex, and API for trusted-access enterprise customers (Amgen, Moderna, Allen Institute, Thermo Fisher).
OpenAI's first purpose-built domain model — frontier reasoning tuned on molecules, proteins, genes, and experimental workflows.
Key specs
| Codex prediction ranking | above 95th percentile of human experts |
|---|---|
| Codex sequence generation | 84th percentile |
What is it?
GPT-Rosalind, announced April 16 and rolling out April 17 in research preview, is OpenAI's first purpose-built domain-specific frontier reasoning model, aimed squarely at life sciences and drug discovery. Named after Rosalind Franklin, it is available through ChatGPT, Codex, and the API to a 'trusted access' program of qualified enterprise customers in the US — launch partners include Amgen, Moderna, the Allen Institute, and Thermo Fisher.
How does it work?
The model is tuned for reasoning over molecules, proteins, genes, pathways, and disease-relevant biology. It supports multi-step scientific workflows: literature review, evidence synthesis, genomics analysis, hypothesis generation, experimental planning, and interpretation of research data. In Codex's internal evaluations, GPT-Rosalind submissions ranked above the 95th percentile of human experts on prediction tasks and reached the 84th percentile on sequence generation. During the preview, the model does not consume existing credits or tokens.
Why does it matter?
Domain-specific frontier models are a different product category from general-purpose GPT-5.x — narrower but deeper. If the Codex prediction numbers hold outside OpenAI's internal eval, Rosalind is the first public frontier model explicitly designed as a lab partner rather than a chatbot. Combined with the parallel Novo Nordisk partnership announced the same week, this is OpenAI signaling a serious enterprise push into pharma.
Who is it for?
Biotech and pharma research teams; computational biologists; drug-discovery infra leads.
Try it
https://openai.com/index/introducing-gpt-rosalind/ (research preview; trusted-access program)