LVBench evaluates long-video understanding across extended videos such as television, sports and surveillance footage. It measures information extraction, temporal grounding, reasoning and summarization over videos whose duration can span hours.
查看 LVBench 的最新得分、模型模式、发布时间与参数规模,快速了解当前完整榜单表现。
数据优先来自官方发布(GitHub、Hugging Face、论文),其次为评测基准官方结果,最后为第三方评测机构数据。 了解数据收集方法
| 排名 | 模型 | 开源情况 | |||
|---|---|---|---|---|---|
![]() Gemini 3.7 Flash 思考水平·中 | 85.40 | 2026-08-13 | 未知 | 闭源 |
Google reported 85.4% for Gemini 3.7 Flash using 1,024 frames and no tools in its official model card on 2026-08-13.