VCIFBench - Format: leaderboard
VCIFBench evaluates complex instruction following in video understanding over 306 constraint-rich instructions spanning content, format, style and structure requirements, judged by a hybrid rule-based, executable and GPT-5 pipeline. This board reports Constraint pass rate (%) on format constraints (serialization, length, fixed patterns); higher is better.
Metric: Constraint pass rate (%) on format constraints (serialization, length, fixed patterns); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 92.49 | #123 |
| 2 | Gemini 3 Pro | 91.39 | #137 |
| 3 | Gemini 3 Flash | 89.36 | #164 |
| 4 | GLM-4.1V-9B (Thinking) | 81.38 | #1060 (GLM-4.1V-9B) |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=vcifbench-format · How It Works · Data refreshed daily, snapshot 2026-11-13.