Benchmark
MRCR v2 · 256K
Scores are percentages. Higher is better.
Benchmark sourceScores from BenchBoard docs.
- Models measured
- 4
- Scores recorded
- 4
- Sources
- 1
- IQ weight
- No weight published
Score distribution
- 89.2%
- Mean
- 92.9%
- Median
- 4.4%
- Std dev
- 83.2%
- 10th percentile
- 93.8%
- 90th percentile
- 83.2%
- Min
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.6 SolQwen表:MRCR v2 · 256K · 8-needle· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 93.8% |
| 2 | Qwen3.8 MaxQwen表:MRCR v2 · 256K · 8-needle· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 92.9% |
| 3 | Qwen3.7 MaxQwen表:MRCR v2 · 256K · 8-needle· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 86.7% |
| 4 | Claude Opus 4.8Qwen表:MRCR v2 · 256K · 8-needle· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 83.2% |