Make the video: the AI base, then your library's b-roll cut in (spends video credits)

Make one video from a written script. The render options match BulkGenerateDTO's.

mode "ai_base": the AI base video, then the library's b-roll cut in. mode "footage_edit": cut from the library, one moment per beat, with AI only for the beats the plan generates, then the voiceover, music and on-screen text. mode "presenter": the plan's character talks on camera in Seedance takes with their own voice; footage, generated scenes and motion graphics play over them, then captions, sound effects and music.