In its October 30 ChatGPT Business release notes, OpenAI said GPT-6 Rosalind Research is rolling out to a limited set of customers with Research approval. It shows up in the model picker in ChatGPT Work and Codex, not in the Chat tab, and separately approved organizations can call it through the API as gpt-6-rosalind. OpenAI's token-based Enterprise rate card lists it at $15 per million input tokens, $1.50 cached input and $75 output, three times the price of GPT-5.5 Rosalind Research.
Key Takeaways
- ✓Approved orgs only: both the organization and the user need Rosalind Research approval, and Discovery approval does not include it
- ✓Where to use it: the model picker in ChatGPT Work and Codex, not the Chat tab; in Codex it pairs with the Life Sciences Research plugin
- ✓API model name gpt-6-rosalind needs separate approval and is limited to internal research tools, not customer-facing products
- ✓Priced at $15 input, $1.50 cached input and $75 output per million tokens, versus $5/$0.50/$25 for GPT-5.5 Rosalind Research
- ✓Enterprise workspaces start with a two-week Admin Preview that is off by default until an admin enables it for approved users
Key Decision Metrics at a Glance
Turn your technical choice into a development budget
Compare 40 dev plans & simulate token costs vs $20/mo subscriptions
Project Links & Resources
Direct AccessIn-Depth Technical Analysis
Background and the problem
Life sciences research mixes literature, genomics and sequencing data, protein sequences, compound data and lab records. General models can write code and summarize papers, but they often fall short on biological reasoning and on working across specialist tools, and biosecurity concerns keep model makers from opening up their strongest biology capabilities to everyone. OpenAI's answer is a separate life sciences brand, Rosalind, which pairs the GPT-Rosalind model family with Rosalind Workbench and gates access through review. GPT-6 Rosalind Research brings that family up to the GPT-6 generation.
Architecture and how it works
According to the OpenAI Help Center, GPT-Rosalind combines specialized biological reasoning with scientific tool use for target research, genomics and sequencing analysis, protein and sequence analysis, medicinal chemistry, literature synthesis, wet-lab troubleshooting and experiment planning. Access comes in two tiers. Research covers frontier biological work and includes GPT-6 Rosalind Research and GPT-5.5 Rosalind Research. Discovery covers everyday scientific work and uses the existing Rosalind 5.5 model with Discovery-specific safeguards and stricter monitoring than Research. Each tier is reviewed separately, so Discovery approval does not grant Research access. In Codex, the Life Sciences Research plugin connects these workflows to scientific tools, databases and data sources.
Benchmarks and measured results
OpenAI's release notes and Help Center article do not publish evaluation results for GPT-6 Rosalind Research, and there is no independent testing yet, so this article cites no scores. What can be compared is price. On OpenAI's token-based Enterprise rate card, GPT-6 Rosalind Research costs $15 per million input tokens, $1.50 cached input and $75 output. GPT-5.5 Rosalind Research and GPT-5.5 Rosalind Discovery are both $5, $0.50 and $25, and the general flagship GPT-6 Astra is $10, $1 and $50. The new model is three times the price of the previous generation and 50% more than GPT-6 Astra.
Practical guide for developers
Teams that already have Research approval can open Work or Codex in the approved workspace and pick GPT-6 Rosalind Research in the model picker. Codex users should install the Life Sciences Research plugin first, though installing it does not grant model access by itself. Enterprise workspaces start with a two-week Admin Preview during which the model is off by default, and existing Research approval does not skip it, so an admin has to enable it for approved users. API users call the model as gpt-6-rosalind, which requires the API organization to be separately approved and is currently limited to internal research tools and workflows, not customer-facing products. Existing GPT-5.5 Rosalind Research workflows keep working, so it makes sense to compare quality and token use on a few representative tasks before switching everything over. U.S. university labs can also apply for the Rosalind Academic Plan, which offers up to 30 complimentary ChatGPT Business Standard seats plus GPT-Rosalind access.
Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.
Discussion & Comments
0Sign in to join the discussion
Connect with AI developers to exchange benchmark insights.