VideoVIBE - Feature Execution: leaderboard
Metric: Failure-diagnosis accuracy (%; questions of the Feature Execution subdimension of Functional Executability; zero-shot single-choice diagnosis from the interaction video and the page source code (Video+HTML), about 1.7K questions built from 6,338 verified failures of one-shot generated web applications recorded under human operation; free-form answers mapped deterministically to an option). Source: arxiv.org. Saturation forecast: Around December 2026. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Flash | 90.37 |
| 2 | Gemini 3.1 Flash Lite | 89.84 |
| 3 | Qwen 3 VL 8B (Thinking) | 89.3 |
| 4 | Qwen 3 VL 30B A3B (Thinking) | 88.24 |
| 5 | Gemini 3.1 Pro (Preview) | 86.1 |
| 6 | Qwen 3 VL 30B A3B Instruct | 86.1 |
| 7 | Qwen 3.5 35B A3B | 83.96 |
| 8 | Gemini 3 Flash | 83.42 |
| 9 | Qwen 3 VL 8B Instruct | 79.68 |
| 10 | GLM-4.6V | 77.01 |
| 11 | MiMo-V2.5 | 73.26 |
Interactive version: theaggregate.ai/benchmark?slug=videovibe-feature-execution · How It Works · Data refreshed daily, snapshot 2026-09-29.