SVCBench - E1-Transit: leaderboard
Metric: Gaussian Precision Accuracy (0-100): per query point exp(-(prediction - truth)^2 / (2 sigma^2)) with sigma 5% of the true count, averaged over the 48 E1-Transit (scene transitions) questions' query points; unparseable answers score 0; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 49.8 | #93 |
| 2 | GPT-5.4 | 42 | #76 |
| 3 | Seed 1.8 | 38.2 | #136 |
| 4 | Kimi K2.5 | 33.5 | #139 |
| 5 | Qwen 3.5 35B A3B | 27.7 | #250 |
| 6 | Qwen 2.5 VL 7B Instruct | 13.3 | #643 |
| 7 | Molmo2-8B | 2.7 | #522 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=svcbench-e1-transit · How It Works · Data refreshed daily, snapshot 2026-10-11.