OpenKatana is a free, open-source video editor that your AI agent drives with code
OpenKatana is a new open-source video editor where an AI coding agent plans the shots and builds the edit, with cuts, music, motion graphics and 3D scenes, while the project stays editable on your own computer. It works from a command line, through MCP, as ComfyUI apps and nodes, or in a local Studio app where you review and fix the result by hand.
What it does
You give your agent a brief plus photos, footage or music; it writes scene and animation code and calls OpenKatana's 72 tools for cuts, speed ramps, transitions, beat analysis, audio mixing, typography and code-built WebGL 3D scenes, with an optional Blender render path. In ComfyUI it adds five saved apps and 82 nodes, so you can connect a video model's output to Import media, Make edit and Render to video, then open the same project in Studio.
Results and limits
The README's sample film, MOTIF, is a 24-second 1080p60 product film with procedural 3D objects and an original soundtrack, made with the Blender option and no video-generation model. A bundled example renders an eight-second edit from included images and music with no agent or model, and a tested LTX-2.5 example pairs a cloud generation with local speed ramp, freeze, typography and audio edits.
Worth knowing
It is version 0.1.0 (released 22 October), built for Comfy's #OpenSourceHiggsfield challenge and inspired by Higgsfield's Katana, but not affiliated with Higgsfield. It needs Python 3.11+, Git, FFmpeg and a supported browser; Blender and AI models are optional. CI covers Windows and Linux, macOS is untested, there is no login, and the code is Apache 2.0.