GPT-Rosalind

openai-gpt-rosalind

fields

nameREQ	GPT-Rosalind	openai.com	2026-04-30
vendorREQ	OpenAI	openai.com	2026-04-30
released_at	2026-04-16	openai.com	2026-04-30
modality	text, code, biology2 revisions	openai.com	2026-05-05
access	closed	openai.com	2026-04-30
context_window
parameters
benchmark_score	75.1	www.technology.org	2026-05-05
benchmark_name	BixBench Pass@1	www.technology.org	2026-05-05
announcement_url	https://openai.com/index/introducing-gpt-rosalind/	openai.com	2026-04-30

history · 8 fields · 9 revisions

name1 revision

GPT-Rosalind current openai.com · 2026-04-30

GPT-Rosalind: frontier reasoning model built to support research across biology, drug discovery, and translational medicine. Research preview via trusted access program.

vendor1 revision

OpenAI current openai.com · 2026-04-30

GPT-Rosalind: frontier reasoning model built to support research across biology, drug discovery, and translational medicine. Research preview via trusted access program.

released_at1 revision

2026-04-16 current openai.com · 2026-04-30

GPT-Rosalind: frontier reasoning model built to support research across biology, drug discovery, and translational medicine. Research preview via trusted access program.

modality2 revisions

text, code, biology current openai.com · 2026-05-05

GPT-Rosalind is a reasoning model tuned specifically for biological research, drug discovery, and translational medicine.

text, code superseded openai.com · 2026-04-30

GPT-Rosalind: frontier reasoning model built to support research across biology, drug discovery, and translational medicine. Research preview via trusted access program.

access1 revision

closed current openai.com · 2026-04-30

GPT-Rosalind: frontier reasoning model built to support research across biology, drug discovery, and translational medicine. Research preview via trusted access program.

benchmark_score1 revision

75.1 current www.technology.org · 2026-05-05

On the BixBench bioinformatics benchmark, the model scored 0.751 Pass@1, ahead of GPT-5.4 (0.732), Grok 4.2 (0.698), GPT-5.2 (0.611).

benchmark_name1 revision

BixBench Pass@1 current www.technology.org · 2026-05-05

On the BixBench bioinformatics benchmark, the model scored 0.751 Pass@1

announcement_url1 revision

https://openai.com/index/introducing-gpt-rosalind/ current openai.com · 2026-04-30

GPT-Rosalind: frontier reasoning model built to support research across biology, drug discovery, and translational medicine. Research preview via trusted access program.