reasoning benchmark
MRCR v2 (8-needle, 512K-1M)
MRCR v2 8-needle variant evaluated on contexts from 512K to 1M tokens.
Updated Aug 7, 2026
Published models3
Registry coverage3
MetricScore
EvidenceB
MRCR v2 (8-needle, 512K-1M) leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | OP | 73.8% | 100.0% | 3 | C | |
| 2 | OP | 72.5% | 50.0% | 3 | C | |
| 3 | OP | 41.3% | 0.0% | 3 | C |
Score distribution
Top published rows on the benchmark's original scale.
MRCR v2 (8-needle, 512K-1M) highlights
The top published results on this benchmark's own scale.
What is MRCR v2 (8-needle, 512K-1M)?
Definition and scoring fields from the benchmark registry.
MRCR v2 8-needle variant evaluated on contexts from 512K to 1M tokens.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MRCR v2 (8-needle, 512K-1M)
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mrcr-v2-8-needle-512k-1m|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MRCR v2 (8-needle, 512K-1M).
Which model scores highest on MRCR v2 (8-needle, 512K-1M)?
GPT-5.6 Sol is currently ranked first with 73.8%.
What does MRCR v2 (8-needle, 512K-1M) measure?
MRCR v2 8-needle variant evaluated on contexts from 512K to 1M tokens.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
3 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.