OpenAI releases GPT-6 Rosalind Research through trusted access

Availability and evidence

OpenAI’s October 30 changelog expressly released GPT-6 Rosalind for internal research by authorized users of approved organizations. The accompanying safety addendum describes a trusted-access model for biology, drug discovery and translational medicine. [1][2]

The addendum says Rosalind uses the same weights as GPT-6 Astra, with no added training data or further post-training. Its distinction is the restricted deployment and safeguards for legitimate biological research, including some dual-use topics, while retaining boundaries against biological weaponization. [2]