STRAND: leaderboard
Metric: Faithful accuracy (%; share of the 977 target yes/no questions over 88 videos for which the model answers the target and every prerequisite sub-question correctly, computed over all targets; 2,516 sub-questions; end-to-end video input; majority-class predictor 19.1). Source: arxiv.org. Saturation forecast: Around 2034. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 VL 32B (Thinking) | 46.1 |
| 2 | Qwen 3.5 27B | 45.5 |
| 3 | Gemini 3.1 Pro (Preview) | 38.9 |
| 4 | Claude Sonnet 4.6 | 30.2 |
| 5 | InternVL3-78B | 29.7 |
| 6 | Gemini 3 Flash | 28.7 |
| 7 | GPT-5 | 20 |
Interactive version: theaggregate.ai/benchmark?slug=strand · How It Works · Data refreshed daily, snapshot 2026-09-26.