GPT Rosalind Research Reaches General Availability
OpenAI’s GPT Rosalind Research is now generally available for approved life sciences research. Billing starts October 5, 2026, with specific token pricing.
General Availability and Access Scope
On September 8, 2026, OpenAI confirmed that the GPT Rosalind Research model, identified as gpt-rosalind-research, is generally available. This release is part of a trusted-access program specifically designated for approved internal life sciences research. The announcement distinguishes this historical event date from the current publication date of this article.
The source establishes that access is restricted to approved entities within the life sciences sector. It does not provide details on the application process, approval criteria, or the specific organizations currently granted access. The model is described as having text input and text output modalities. No context window size or maximum output token limit is specified in the available facts.
Pricing Structure and Billing Start Date
Standard pricing for GPT Rosalind Research is explicitly stated in the official changelog. The cost structure includes $5 per 1 million input tokens, $0.50 per 1 million cached input tokens, and $25 per 1 million output tokens. These figures represent the standard rates provided by the source.
Billing for this model begins on October 5, 2026. This date marks the start of chargeable usage under the general availability status. The source does not mention free tiers, promotional periods, or volume discounts. It also does not clarify how cached input tokens are generated or managed within the system.
Evaluation Procedure and Source Limitations
Engineers evaluating this model should design tests based on their specific life sciences research tasks. A practical procedure involves creating a standardized dataset of relevant queries. Measure the model's responses against predefined accuracy metrics. Compare these results with existing baselines used in your organization. Validate that the output format meets downstream processing requirements.
The source confirms only the general availability status, the target audience, the modality types, and the pricing schedule. It does not confirm performance benchmarks, response speed, architectural details, or local installation capabilities. The fact that open weights are not confirmed does not establish that local installation is impossible or that API access is mandatory. Any claims about improved efficiency or specific technical advantages remain unverified by the provided facts.
- Define test cases relevant to internal life sciences research.
- Measure token consumption to align with the $5/$0.50/$25 pricing tiers.
- Compare output quality against current internal tools.
- Validate that the trusted-access approval covers your specific use case.
Primary sources
Frequently asked questions
When does billing for GPT Rosalind Research begin?
Billing begins on October 5, 2026. Prior to this date, the model was not subject to standard billing under the general availability terms described in the source.
What are the standard pricing rates for the model?
The standard pricing is $5 per 1 million input tokens, $0.50 per 1 million cached input tokens, and $25 per 1 million output tokens. These rates apply to general availability usage through the trusted-access program.