Introducing new capabilities to GPT-Rosalind

· Source: OpenAI News · Field: Science & Research — Artificial Intelligence & Machine Learning, Life Sciences & Biology, Health & Medical Research · Depth: Expert, long

Summary

OpenAI has released an update to its GPT-Rosalind series, a model specifically designed for enterprise-scale life sciences research. This new version integrates GPT-5.5's agentic coding and tool-use capabilities with enhanced intelligence in critical drug-discovery areas like medicinal chemistry and genomics. The updated GPT-Rosalind demonstrates broad performance gains across various research tasks, including complex medicinal chemistry queries, quantitative biology, and wet lab troubleshooting. It leads performance on the new LifeSciBench, an expert-judged benchmark covering six workflow areas. Specifically, it outperforms GPT-5.5 on MedChemBench (27.5% vs. 25.1%, 7.2% fewer tokens), GeneBench (21.6% vs. 20.4%, 31% fewer tokens), and LabWorkBench (63.2% vs. 55.8%, 5.3% fewer tokens). Access is expanding globally through a trusted-access deployment structure, with Novo Nordisk already leveraging its capabilities.

Key takeaway

For AI/ML Directors in life sciences seeking to accelerate drug discovery and research, GPT-Rosalind offers a specialized solution. You should evaluate its enhanced agentic capabilities and domain-specific performance in medicinal chemistry and genomics. Consider requesting access to its research preview to integrate its plugins for evidence retrieval and bioinformatics execution into your existing workflows, potentially streamlining complex analyses and improving decision-making.

Key insights

GPT-Rosalind integrates advanced AI with scientific workflows, significantly improving life sciences research efficiency and accuracy.

Principles

Method

GPT-Rosalind uses agentic coding and tool-use with specialized plugins (Life Sciences Research, Life Sciences NGS Analysis) to integrate evidence retrieval, biological interpretation, and bioinformatics execution within a unified workspace like Codex.

In practice

Topics

Code references

Best for: AI Engineer, Machine Learning Engineer, CTO, Research Scientist, AI Scientist, Director of AI/ML

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by OpenAI News.