GPT-Rosalind Leaves Research Preview for Eligible Organizations Worldwide
GPT-Rosalind, OpenAI's life sciences reasoning model, is out of research preview for eligible organizations worldwide through OpenAI's trusted-access program. OpenAI says qualified customers can use it in the API, Codex, and ChatGPT Enterprise, and a Life Sciences research plugin connects Codex to 50-plus public tools. Billing began October 5, 2026: $5 input and $25 output per million API tokens.
OpenAI says GPT-Rosalind is leaving research preview and is now available globally to eligible organizations through its trusted-access program. The life sciences model is available to qualified customers through the API, Codex, and ChatGPT Enterprise, with the latest Rosalind models included as OpenAI releases them. (OpenAI Developers on X; OpenAI's GPT-Rosalind announcement)
The change is an access and maturity update rather than a new model family. GPT-Rosalind was introduced in April as a purpose-built reasoning model for biology, drug discovery, and translational medicine; OpenAI's September 11 update changes its status from research preview to trusted access for eligible organizations. (OpenAI's GPT-Rosalind announcement)
Key Takeaways
- GPT-Rosalind is out of research preview for eligible organizations worldwide.
- Access is gated. OpenAI describes a trusted-access program, not unrestricted self-serve availability.
- The listed surfaces are the API, Codex, and ChatGPT Enterprise.
- A Codex Life Sciences research plugin is available more broadly, with skills for common scientific workflows and connections to more than 50 public tools and data sources.
- Billing began October 5, 2026. OpenAI's API pricing page lists
gpt-rosalind-researchat $5.00 input, $0.50 cached input, and $25.00 output per million tokens.
What GPT-Rosalind Is Designed to Do
GPT-Rosalind is optimized for scientific workflows across chemistry, protein engineering, genomics, and translational research. OpenAI describes the model as combining deeper scientific understanding with improved tool use so researchers can synthesize evidence, generate hypotheses, plan experiments, and analyze data across long, multi-step investigations. (OpenAI's GPT-Rosalind announcement)
The developer-facing announcement describes the practical workflow in similar terms: connect findings across papers and experimental results, weigh evidence for a biological target, and work through an analysis that helps plan what to test next. (OpenAI Developers on X)
This is a domain model, not a claim that GPT-Rosalind replaces scientific review. OpenAI frames it as a system for helping researchers move from evidence to hypotheses and defensible decisions, while trusted access provides the governance layer for organizations using it in professional research environments. (OpenAI's GPT-Rosalind announcement)
Where the Model Is Available
| Surface | What OpenAI says |
|---|---|
| API | Eligible organizations can access GPT-Rosalind through OpenAI's API |
| Codex | Researchers can use GPT-Rosalind in Codex, including with Life Sciences workflows |
| ChatGPT Enterprise | Qualified customers can access the model in ChatGPT Enterprise |
The September update says the trusted-access program is global for eligible organizations. It also says that eligible organizations will continue to receive access to new GPT-Rosalind models as they are released. (OpenAI Developers on X; OpenAI's GPT-Rosalind announcement)
OpenAI does not describe this as an open, no-review rollout. The original announcement says organizations are evaluated for legitimate scientific use, governance and safety oversight, misuse-prevention controls, and controlled access in secure environments. (OpenAI's GPT-Rosalind announcement)
The Life Sciences Plugin Makes Codex More Useful
The model update is paired with a Codex package for researchers. OpenAI says its Life Sciences research plugin is freely accessible and includes modular skills for workflows such as human genetics, functional genomics, protein structure, biochemistry, clinical evidence, and public study discovery. (OpenAI's GPT-Rosalind announcement; OpenAI Developers on X)
The plugin connects to more than 50 public multi-omics databases, literature sources, and biology tools. Example workflows include looking up protein structures, searching sequences, reviewing literature, and discovering public datasets. OpenAI says all users can use the plugin package with its mainline models, while eligible Enterprise users can pair it with GPT-Rosalind for deeper biological reasoning. (OpenAI's GPT-Rosalind announcement)
That division matters for teams deciding what to adopt. The research plugin is the broadly usable workflow layer; GPT-Rosalind remains the specialized model gated by trusted access and organizational qualification.
What the Preview Exit Means for Teams
The end of research preview gives qualified organizations a clearer path to evaluate GPT-Rosalind as part of a governed research stack. Teams can assess the model in the surface they already use -- API, Codex, or ChatGPT Enterprise -- then decide whether the plugin's database and literature workflows fit their internal process. (OpenAI's GPT-Rosalind announcement)
The access model also sets expectations: organizations should prepare qualification and safety-review information rather than assume an instant model toggle. OpenAI's description emphasizes beneficial use, strong governance, controlled access, and enterprise-grade security. (OpenAI's GPT-Rosalind announcement)
Pricing
Billing for gpt-rosalind-research began on October 5, 2026, the date OpenAI's September 11 update gave for its published pricing (OpenAI's GPT-Rosalind announcement). OpenAI's pricing pages now list these rates per million tokens:
| Surface | Input | Cached input | Output |
|---|---|---|---|
OpenAI API (gpt-rosalind-research) | $5.00 | $0.50 | $25.00 |
| Codex credits (GPT-Rosalind-Research) | 125 credits | 12.5 credits | 625 credits |
In the API, cache-write pricing does not apply to this model, access is limited to approved internal research through the trusted-access program, and regional processing (data residency) endpoints carry a 10% uplift for eligible models released on or after March 5, 2026 (OpenAI API pricing). The Codex rates are Standard-speed credits, and Codex credit billing has no separate cache-write charge (ChatGPT docs: Pricing).
Bottom Line
GPT-Rosalind has moved from research preview into trusted access for eligible organizations worldwide. The practical launch surface is the combination of a specialized life sciences model in the API, Codex, and ChatGPT Enterprise, plus a more broadly usable Codex research plugin that connects mainline models to public scientific tools and data. Teams interested in the model should start with the qualification process; billing has applied since October 5, 2026, at $5 input and $25 output per million API tokens. (OpenAI's GPT-Rosalind announcement; OpenAI API pricing)
Corrections and Updates
- October 9, 2026: This page said published pricing takes effect October 5, 2026, without a price; billing has applied since that date, and OpenAI lists
gpt-rosalind-researchat $5.00 input, $0.50 cached input, and $25.00 output per million tokens.
Sources
- OpenAI Developers on X, “GPT-Rosalind is out of research preview for eligible organizations worldwide,” September 11, 2026: https://x.com/OpenAIDevs/status/2098509255947293035
- OpenAI Developers on X, “Bring stronger biological reasoning to your research with GPT-Rosalind,” September 11, 2026: https://x.com/OpenAIDevs/status/2098509226335412479
- OpenAI Developers on X, “In Codex, use Life Sciences plugins,” September 11, 2026: https://x.com/OpenAIDevs/status/2098509243389485252
- OpenAI, “Introducing GPT-Rosalind for life sciences research”: https://openai.com/index/introducing-gpt-rosalind/
- OpenAI API pricing (Life Sciences: gpt-rosalind-research): https://developers.openai.com/api/docs/pricing
- ChatGPT docs, Pricing (Codex token rates): https://learn.chatgpt.com/docs/pricing
Read next
Related coverage- OpenAI
OpenAI Opens Its Codex-Powered Agents API to All Developers in Public Beta
OpenAI's Agents API is now in public beta for all developers. The managed service uses the Codex harness to run long-lived cloud agents, coordinate parallel subagents, compact context across long sessions, and work in OpenAI-hosted, VPC, or partner-provided sandboxes.
- OpenAI
OpenAI Releases Codex Security CLI for Repository Scans and CI Checks
OpenAI says its open-source Codex Security CLI can scan repositories, track findings across runs, verify fixes, and add security checks to CI/CD. The beta CLI requires Codex Security access and is built for teams that want code-aware security review in the terminal.
- OpenAI
OpenAI Launches ChatGPT for Financial Services
OpenAI introduced ChatGPT for Financial Services, a tailored ChatGPT Work experience for eligible financial institutions. It combines hosted datasets from providers such as Daloopa, PitchBook, LSEG News, and Crunchbase with GPT-6 Astra, granular citations, spreadsheet and presentation generation, firm templates, and enterprise access controls.
Frequently Asked Questions
What changed for GPT-Rosalind?
OpenAI said on September 11, 2026 that GPT-Rosalind, its life sciences reasoning model introduced in April, is leaving research preview and is available globally to eligible organizations through its trusted-access program, which evaluates organizations before granting access.
Where can eligible organizations use GPT-Rosalind?
OpenAI lists the API, Codex, and ChatGPT Enterprise as access surfaces for qualified customers. Eligible organizations continue to get access to new GPT-Rosalind models as OpenAI releases them.
What is the Codex Life Sciences research plugin?
It is a freely accessible Codex plugin package with modular skills that connect models to more than 50 public multi-omics databases, literature sources, and biology tools. All users can run it with mainline models; eligible Enterprise users can pair it with GPT-Rosalind.
How much does GPT-Rosalind cost?
Billing for gpt-rosalind-research began October 5, 2026. OpenAI's API pricing page lists $5.00 input, $0.50 cached input, and $25.00 output per million tokens, with no cache-write charge. In Codex, it costs 125, 12.5, and 625 credits per million input, cached input, and output tokens.