OpenAI’s GPT-Rosalind API will begin billing on Oct. 5 at $5 per million input tokens, with output priced at $25 per million tokens. Cached input costs $0.50 per million. The rates apply to a specialized life-sciences model that remains available only to organizations approved for trusted access.
The API pricing page lists the model as gpt-rosalind-research. OpenAI describes GPT-Rosalind as a model for research across biology, drug discovery and translational medicine, with tool use for scientific workflows. The company first announced it in April and said in a September update that eligible organizations could access it globally through its trusted-access program.
The access boundary still matters: OpenAI says API use is limited to approved internal research tools, workflows and applications. The model is not available through the API for customer-facing products or external commercial applications. Eligible organizations need Enterprise or Business agreements and must receive approval through the trusted-access program.
Sources
- OpenAI API pricing
- GPT-Rosalind access and API restrictions
- OpenAI’s GPT-Rosalind announcement and September update
Image: Protein Ribbon, courtesy of NIAID / Ryan Kissinger, via Wikimedia Commons; public domain in the United States.
.png){kind=link}