In long-video understanding, conventional uniform frame sampling often fails to capture key visual evidence, leading to degraded performance and increased hallucinations. To address this, recent agentic thinking-with-videos paradigms have emerged, adopting a localize-clip-answer pipeline in which the model actively identifies relevant video segments, performs dense sampling within those clips, and then produces answers. However, existing methods remain inefficient, suffer from weak localization, and adhere to rigid workflows. To solve these issues, we propose VideoTemp-o3, a unified agentic thinking-with-videos framework that jointly models video grounding and question answering. VideoTemp-o3 exhibits strong localization capability, supports on-demand clipping, and can refine inaccurate localizations. Specifically, in the supervised fine-tuning stage, we design a unified masking mechanism that encourages exploration while preventing noise. For reinforcement learning, we introduce dedicated rewards to mitigate reward hacking. Besides, from the data perspective, we develop an effective pipeline to construct high-quality long video grounded QA data, along with a corresponding benchmark for systematic evaluation across various video durations. Experimental results demonstrate that our method achieves remarkable performance on both long video understanding and grounding.

翻译：在长视频理解任务中，传统的均匀帧采样方法往往难以捕捉关键的视觉证据，导致性能下降和幻觉现象加剧。为解决这一问题，近期出现了基于智能思考的视频理解范式，采用“定位-剪辑-回答”的流程，即模型主动识别相关的视频片段，在这些片段内进行密集采样，然后生成答案。然而，现有方法效率仍然较低，存在定位能力弱的问题，且遵循僵化的工作流程。为解决这些不足，我们提出了VideoTemp-o3，一个统一的基于智能思考的视频理解框架，它联合建模视频时序定位与问答任务。VideoTemp-o3展现出强大的定位能力，支持按需剪辑，并能修正不准确的定位结果。具体而言，在有监督微调阶段，我们设计了一种统一的掩码机制，以鼓励探索同时防止噪声干扰。在强化学习阶段，我们引入了专门的奖励函数以缓解奖励欺骗问题。此外，从数据角度出发，我们开发了一个高效的流水线来构建高质量的长视频时序定位问答数据，并建立了一个相应的基准测试集，用于系统评估不同视频时长下的性能。实验结果表明，我们的方法在长视频理解和时序定位任务上均取得了卓越的性能。