COL(SZ:300364) Xin Lang Cai Jing·2026-02-10 06:44Group 1 - ByteDance's Seedance 2.0 video model is currently in internal testing, with the Xiaoyunke app available for experience, and Seedance 2.0 fast is set to launch soon [1] - Seedance 2.0 supports the generation of single-segment videos lasting 5 to 15 seconds, utilizing a self-developed storyboard workflow to create content with multiple angles, character dialogues, and subtitles [1] - As of February 10, 2026, the Invesco AI ETF (159142) has risen by 1.53%, with intraday highs nearing 2%, and key component stocks such as Zhongwen Online and Guotou Intelligent showing significant gains [1] Group 2 - The video generation sector is entering a "dashboard era" of precise control, with domestic multimodal models like Lingxi 3.0 and Jimeng 2.0 significantly accelerating iteration [2] - The latest versions of these models support multimodal input of images, audio, and video, enhancing instruction adherence, storyboard controllability, and secondary editing capabilities [2] - Single generation duration has been extended to 15 seconds, with basic metrics such as physical realism and motion smoothness approaching international leading levels [2]