Introducing new capabilities to GPT-Rosalind
OpenAI Blog · 2026-06-03
OpenAI announces an updated GPT-Rosalind model series purpose-built for enterprise life sciences research, combining GPT-5.5's agentic tool-use with stronger performance in drug discovery, medicinal chemistry, and genomics, outperforming GPT-5.5 on three new domain-specific benchmarks while using fewer tokens.
Appears in
Extraction
Topics: life-sciences-aidrug-discoverygenomicsai-benchmarksbiomedical-ai
Claims
- Updated GPT-Rosalind combines GPT-5.5 agentic coding and tool-use capabilities with deeper performance in medicinal chemistry and genomics workflows.
- GPT-Rosalind outperforms GPT-5.5 on MedChemBench (27.5% vs 25.1%) while using 7.2% fewer tokens.
- On GeneBench for long-horizon genomics analysis, GPT-Rosalind achieves higher accuracy (21.6% vs 20.4%) using 31% fewer tokens than GPT-5.5.
- GPT-Rosalind scores 63.2% on LabWorkBench for wet lab troubleshooting versus GPT-5.5's 55.8%, using 5.3% fewer tokens.
- Novo Nordisk is partnering with OpenAI to use GPT-Rosalind to accelerate drug discovery and medical research at scale.
Key quotes
Life sciences research is complex, data-rich, and interdisciplinary. To deliver meaningful value for researchers, advanced AI models must be grounded in trusted scientific data, connected to validated tools, and integrated into the real-world workflows researchers use every day.
GPT-Rosalind is now available in research preview to eligible organizations globally through our trusted-access deployment structure.
This also means applying life sciences AI to high-impact public-benefit work, from drug discovery and translational medicine to public health, preparedness, and biodefense.