PredicateLongBench - Clustered Near-Sorted Decoys (LongBench): leaderboard
Metric: Accuracy (%; share of the 93 LongBench v2 documents under 170K tokens used as word lists whose answer is exactly the one word sequence satisfying the predicate, locate the sorted run with 8 near-sorted decoys clustered next to it; up to 16K output tokens including reasoning, an answer past the cap counts as wrong). Source: arxiv.org. Saturation forecast: Around December 2026. 9 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 (Thinking) | 93 |
| 2 | GPT-5.4 (High) | 68 |
| 3 | Gemini 3.1 Pro (Preview) (High) | 56 |
| 4 | Gemini 3.1 Pro (Preview) (Low) | 50 |
| 5 | Claude Opus 4.6 | 35 |
| 6 | GLM-5.1 | 12 |
| 7 | GPT-5.4 (Non-reasoning) | 4 |
| 8 | Qwen 3.5 397B A17B | 0 |
| 9 | MiniMax-M2.7 | 0 |
Interactive version: theaggregate.ai/benchmark?slug=predicatelongbench-clustered-near-sorted-decoys-longbench · How It Works · Data refreshed daily, snapshot 2026-09-29.