The AI arena is free today

Open Superagent
OpenAIReleased on Sep 11, 2026

GPT-Rosalind: Benchmarks, Pricing & Context Window

GPT-Rosalind is a language model from OpenAI, released in September 2026, with multimodal input, and pricing from $5.00/M input, $0.500/M cached input, $25.00/M output.

GPT-Rosalind is OpenAI's specialized life-sciences reasoning model for approved internal research. OpenAI introduced GPT-Rosalind in research preview on April 16, 2026 and updated the same launch materials on September 11, 2026 to mark

Input
TextImage
Output
Text

GPT-Rosalind benchmarks

Capability tiers

Standing within each category, adjusted for leaderboard depth.

Real tasks performance

High-confidence performance for GPT-Rosalind across real-world prompt categories. Only 95% intervals at most 4 points wide are shown.

Performance by conversation depth

How GPT-Rosalind holds up as conversations get longer.

Quality Tracker

GPT-Rosalind Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

LLM Stats Logollm-stats.com - Wed Oct 07 2026
Notice missing or incorrect data?

GPT-Rosalind pricing

Providers

GPT-Rosalind starts at $5.00 per million input tokens and $25.00 per million output tokens via OpenAI. Reused prompt prefixes cost $0.500 per million cached input tokens.

ProviderInput $/MCached input $/MOutput $/MContext in / outTTFT p95 sOutput p5 c/sModalities in / out
OpenAI logoOpenAI
$5.00$0.500$25.00—/—
—
—
/

Cached input is the discounted price for prompt tokens served from a provider cache. TTFT is time to first token. Output is characters per second; p5 is the sustained floor exceeded by 95% of observed requests.

Try now

huggle
GPT-Rosalindin Huggle

Make it with
GPT-Rosalind.

GPT-Rosalind

GPT-Rosalind latency

GPT-Rosalind time to first token, sustained output throughput, and failed-request rate from live model usage over the trailing 7 days.

GPT-Rosalind examples

Recent arena outputs from GPT-Rosalind, picked from the highest-ranked matchups.

GPT-Rosalind license

GPT-Rosalind is a proprietary model available under its provider's product and API terms.

License
Proprietary
Hosted access

Proprietary license - usage restrictions apply

GPT-Rosalind resources

Official sources for GPT-Rosalind: provider documentation, paper or system card, official launch post.

GPT-Rosalind vs other models

The most-compared alternatives to GPT-Rosalind are GPT-6 Astra, GPT-5.5, GPT-5.6 Sol. Open any pair side-by-side for benchmarks, pricing, context, and latency.

Models like GPT-Rosalind

Models ranked just above and below GPT-Rosalind by LLM Stats score.

 

GPT-6 Astra

Score pending
 

GPT-5.5

Score pending
 

GPT-5.6 Sol

Score pending
 

GPT-5.5 Pro

Score pending
 

GPT-5.6 Terra

Score pending
 

Grok 4.7

Score pending

FAQ

Common questions about GPT-Rosalind.

When was GPT-Rosalind released?

GPT-Rosalind was released on September 11, 2026 by OpenAI. This is the official GPT-Rosalind release date tracked on LLM Stats.

How much does GPT-Rosalind cost?

GPT-Rosalind pricing starts at $5.00 per million input tokens, $0.50 per million cached input tokens, $25.00 per million output tokens via OpenAI, the lowest price among tracked providers.

Who created GPT-Rosalind?

GPT-Rosalind was created by OpenAI.

What is the license for GPT-Rosalind?

GPT-Rosalind is released under the Proprietary license.

Is GPT-Rosalind multimodal?

Yes, GPT-Rosalind is multimodal and can accept both text and images as input.

Where can I use GPT-Rosalind?

GPT-Rosalind is available through 1 provider including OpenAI.

Where is the GPT-Rosalind paper or technical report?

GPT-Rosalind has a paper or technical report available at https://deploymentsafety.openai.com/gpt-rosalind-5-5. Use that source for architecture, training, release and evaluation details.

What models should I compare GPT-Rosalind against?

Common GPT-Rosalind comparisons include GPT-Rosalind vs GPT-6 Astra, GPT-Rosalind vs GPT-5.5, GPT-Rosalind vs GPT-5.6 Sol. Compare them side by side for benchmark scores, pricing, context window, latency and provider availability.