SVHalluc - Global Semantic Alignment: leaderboard
Metric: Accuracy (%; 444 option-balanced video questions, whether the speech describes the visual events (yes/no)). Source: arxiv.org. Saturation forecast: Around December 2026. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 93.1 |
Interactive version: theaggregate.ai/benchmark?slug=svhalluc-global-semantic-alignment · How It Works · Data refreshed daily, snapshot 2026-09-26.