BoundaryBench (NIST High): leaderboard
Metric: Success Rate under NIST high policy (%). Source: boundarybench.com. 14 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Grok 4.5 | 74.9 |
| 2 | GPT-5.6 Sol | 74.2 |
| 3 | Claude Fable 5 | 67.8 |
| 4 | Gemini 3.7 Flash | 66.7 |
| 5 | Claude Opus 5 | 65.2 |
| 6 | Claude Opus 4.8 | 63.3 |
| 7 | GPT-5.6 Terra | 63.3 |
| 8 | Muse Spark 1.2 | 60.7 |
| 9 | GLM-5.2 | 55.8 |
| 10 | Claude Sonnet 5 | 55.1 |
| 11 | Kimi K3 | 54.7 |
| 12 | GPT-5.6 Luna | 53.9 |
| 13 | MiniMax-M3 | 44.2 |
| 14 | Qwen 3.7 Max | 40.4 |
Interactive version: theaggregate.ai/benchmark?slug=boundarybench-nist-high · How It Works · Data refreshed daily, snapshot 2026-09-21.