The specifics:
O1 can process up to seven inputs at simultaneously, analyze text, photos, videos, and subjects, and produce results in three to ten seconds.
Text instructions such as "remove bystanders" or "shift to nighttime" allow users to alter current footage while keeping actors and scenes.
Additional features include start and end frames, multi-subject capabilities, picture, element, action, camera movement, and video references.
According to Kling's internal tests, the model outperforms Runway's Aleph and Google Veo 3.1 in video reference and editing tasks.
Text instructions such as "remove bystanders" or "shift to nighttime" allow users to alter current footage while keeping actors and scenes.
Additional features include start and end frames, multi-subject capabilities, picture, element, action, camera movement, and video references.
According to Kling's internal tests, the model outperforms Runway's Aleph and Google Veo 3.1 in video reference and editing tasks.
December is beginning with some significant AI video advancements thanks to Runway and Kling. Similar to Runway's previous Aleph drop, O1's all-in-one and edit-anything features are enabling granular video alterations like never before, a breakthrough similar to what Nano Banana provided to photos earlier this year.