GUIDE - Help Content Prediction: leaderboard
Metric: Accuracy (%) of choosing the most helpful assistance among four options, with no behavior or intent context, 32 frames per screen-recording segment, video only (no narration), offline setting; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 8 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 55 | #138 |
| 2 | Gemini 2.5 Pro | 52.74 | #145 |
| 3 | Gemini 2.5 Flash | 49.53 | #237 |
| 4 | GPT-4o | 45.95 | #333 |
| 5 | GPT-4o Mini | 31.32 | #588 |
| 6 | InternVL3-8B | 27.03 | #606 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=guide-help-content-prediction · How It Works · Data refreshed daily, snapshot 2026-10-11.