VCIFBench - Structure: leaderboard
Metric: Constraint pass rate (%) on structure constraints (chronological, causal, hierarchical ordering); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 95.24 |
| 2 | Gemini 3 Flash | 93.45 |
| 3 | Gemini 3 Pro | 92.86 |
| 4 | GLM-4.1V-9B (Thinking) | 77.38 |
| 5 | MiniCPM-o-4.5 | 76.79 |
| 6 | InternVL3.5-8B | 49.4 |
Interactive version: theaggregate.ai/benchmark?slug=vcifbench-structure · How It Works · Data refreshed daily, snapshot 2026-09-29.