OpenAI MRCR v2
Saturated
Long-context retrieval benchmark from OpenAI: find and reproduce multiple planted needles across a very long context.
- Domain
- long-context
- Published
- Sep 2026
- Notable for
- Long-context benchmark used at the GPT-6 Astra launch.
Cite
Notes
Only stored in your browser.
Top score 96.3% by GPT-6 Astra - 1 model reporting (1 frontier)
Top models
1
FAQ
- What is OpenAI MRCR v2?
- Long-context retrieval benchmark from OpenAI: find and reproduce multiple planted needles across a very long context.
- What is the current top score on OpenAI MRCR v2?
- The top reported score is 96.3% by GPT-6 Astra, across 1 model reporting (1 from frontier labs).