For the complete documentation index, see llms.txt. This page is also available as Markdown.

Video Models

BonzAI Desktop supports local text-to-video and image-to-video generation.

Video is hardware-intensive. Begin with a short, low-resolution shot and one clear action. Increase quality after movement and composition are correct.

BonzAI+ Live is a different feature: it understands an existing screen/video and creates timestamped cues. Those cues can become generation prompts for recreating or transforming the story in Desktop.

Last updated