OpenAI-MRCR: 2 needle 1M: leaderboard
OpenAI two-needle MRCR task at million-token scale, testing coreference and retrieval under extreme context length.
Metric: Score (%). Source: openai.com. Source: llm-stats.com. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MiniMax M1 40k | 58.6 | #683 |
| 2 | MiniMax M1 80k | 56.2 | #432 |
| 3 | GPT-4.1 | 46.3 | #248 |
| 4 | GPT-4.1 Mini | 33.3 | #355 |
| 5 | GPT-4.1 Nano | 12 | #731 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openai-mrcr-2-needle-1m · How It Works · Data refreshed daily, snapshot 2026-10-20.