Overview
GPT-6 Rosalind is the second model in OpenAI's GPT-Rosalind line of life sciences models, released on October 30, 2026. OpenAI's API changelog describes it as bringing specialized biological reasoning together with scientific tool use for complex life sciences research. It follows GPT-Rosalind, which OpenAI introduced in April 2026 and made generally available to trusted-access customers as gpt-rosalind-research in September 2026.
Access is gated. The model, served under the API identifier gpt-6-rosalind, is available only to organizations approved for Rosalind Research through OpenAI's trusted-access program, for internal research by authorized users. There is no open self-serve signup.
OpenAI's standard price for prompts up to 272K input tokens is $15.00 per 1M input tokens, $1.50 per 1M cached input tokens and $75.00 per 1M output tokens, three times the $5 / $25 rate of gpt-rosalind-research. Longer prompts are billed at $30.00 input, $3.00 cached input and $112.50 output. OpenAI had not published a benchmark table or model card for GPT-6 Rosalind in its developer documentation at release.
| Released | 2026-10-30 |
|---|---|
| License | Proprietary |
| Weights | API only |
| Modalities | Text |
| Status | Available to approved organizations (trusted-access program) |
Pricing
| Input | $15.00 / 1M tokens |
|---|---|
| Cached input | $1.50 / 1M tokens |
| Output | $75.00 / 1M tokens |
Standard tier, prompts up to 272K input tokens. Above 272K: $30.00 input, $3.00 cached input, $112.50 output. Access limited to approved internal research through the trusted-access program.
Strengths
- Specialized biological reasoning combined with scientific tool use, per OpenAI's API changelog
- Served through the API as gpt-6-rosalind for approved Rosalind Research organizations
- Prompt caching at $1.50 per 1M cached input tokens for short-context requests
- Trusted-access deployment limited to authorized users doing internal research
Best for
- Reach for it for complex life sciences research workflows that call scientific tools, inside an organization approved for Rosalind Research.
- Reach for it when an approved lab needs a newer-generation model than gpt-rosalind-research and can absorb the higher token price.
- Reach for gpt-rosalind-research instead when cost matters more, at $5 input and $25 output per 1M tokens.
How to access
| Provider | Model ID |
|---|---|
| OpenAI API — trusted-access program ↗ | gpt-6-rosalind |
GPT-Rosalind — every version
The full lineage of the GPT-Rosalind line, newest first. Every version has its own page — click any to compare specs, benchmarks and pricing.
| Version | Released | Context | License |
|---|---|---|---|
| GPT-6 Rosalindcurrent | 2026-10-30 | — | Proprietary |
| GPT-Rosalind | 2026-04-16 | — | Proprietary |
FAQ
What is GPT-6 Rosalind?
GPT-6 Rosalind is an OpenAI model for life sciences research, released on October 30, 2026. OpenAI describes it as combining specialized biological reasoning with scientific tool use. It is the second model in the GPT-Rosalind line, after the GPT-Rosalind model introduced in April 2026.
Who can use GPT-6 Rosalind?
Only organizations approved for Rosalind Research through OpenAI's trusted-access program, for internal research by authorized users. Approved organizations call it in the API as gpt-6-rosalind.
How much does GPT-6 Rosalind cost?
At OpenAI's standard rate, prompts up to 272K input tokens cost $15.00 per 1M input tokens, $1.50 per 1M cached input tokens and $75.00 per 1M output tokens. Longer prompts cost $30.00 input, $3.00 cached input and $112.50 output per 1M tokens. The earlier gpt-rosalind-research model costs $5 input and $25 output.