IFHierBench: leaderboard
Metric: Prompt-level strict accuracy (%; share of prompts whose response satisfies every constraint in the tree; prompts whose output constraints are nested in a tree (artifact, sections, fields) of depth 0 to 3, checked by Python verifiers as in IFEval; 10 structure-level and 25 content-level constraint types). Source: arxiv.org. Saturation forecast: Around April 2027. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 (High) | 53.7 |
| 2 | Claude Opus 4.6 (High) | 42.8 |
| 3 | Gemini 3 Flash (Preview) | 33.2 |
| 4 | DeepSeek R1 | 15.5 |
| 5 | Kimi K2.5 | 14.7 |
Interactive version: theaggregate.ai/benchmark?slug=ifhierbench · How It Works · Data refreshed daily, snapshot 2026-09-29.