On OpenAI's API pricing page, a name that should not have appeared has recently emerged. A new life sciences model called gpt-rosalind-discovery, which has not yet been officially released, has quietly appeared. Its pricing is exactly the same as that of its well-known sibling model gpt-rosalind-research. Based on this, the outside world has speculated that this new model is aimed at the drug discovery process.

This is not an isolated listing. OpenAI has already established a separate brand for life sciences called Rosalind and has set up multiple product delivery channels - it has previously collaborated with several pharmaceutical companies to explore AI drug development and even launched a bio-defense program. More notably, the identity is: among leading general large model vendors, OpenAI is the first company to establish an independent brand, independent access review, and iteration models specifically for the life sciences field. Its positioning is very restrained, only acting as a tool supplier - selling specialized models to institutions that pass the review, but not getting involved in drug pipelines themselves, never planting seeds in the fields of pharmaceutical companies.

The field of life sciences is extremely tempting for large model vendors: strong payment ability, clear pain points, and a typical high-value vertical scenario. OpenAI's current strategy is to use low prices to attract top pharmaceutical companies and research institutions, and to accumulate data and methods in real scenarios. However, the reality remains: the entire AI4Bio field still lacks large-scale clinical validation, and the role of the model is mostly limited to reducing manual labor, far from being able to independently decide the fate of new drugs.

OpenAI is trying to break down the model types according to tasks in the life sciences, using this detailed approach to quickly monetize the investment. But the current awkward situation is that the differences between these two models at the same price and the actual benefits brought by the new model are still unclear. Whether it can truly stand firm in laboratories and pharmaceutical factories will depend on evaluation and experimental evidence.