Jianying Headless Automates ByteDance Video Drafts

The release of jianying-headless, a source-available Python controller, allows developers and AI coding agents to programmatically generate native draft files for ByteDance's Jianying video editor. Jianying is the Chinese desktop counterpart to CapCut. Instead of relying on fragile graphical user interface automation, this tool directly manipulates the editor's underlying project files, which store timelines, clips, captions, audio tracks, and effects.

The tool supports complex editing features including multi-track timelines, picture-in-picture, subtitles, volume adjustments, and local background music. It exports MP4 files using Jianying's native engine within an isolated process. Currently, the project is pinned to Apple Silicon macOS 26.0+ and Jianying Pro 11.4.2, featuring hash-verified builds and protection against automatic downgrades. The repository is split into two main components: "engine/" for constructing drafts and coordinating exports, and "bridge/" for providing file and pipe adapters to the editor's installed libraries.

The open-source community quickly embraced the project, which amassed over 700 stars on GitHub within days. This surge in popularity occurred after the author reversed initial plans to sell the software for approximately 888 RMB. The controller is currently distributed under a personal, non-commercial license, meaning commercial applications require explicit written authorization.

For video editors and developers, this development enables highly efficient "human-in-the-loop" workflows. Coding agents like Codex can automatically assemble initial video drafts from structured JSON plans. Because the output remains a native project file, human editors can easily open the draft in the desktop application to manually review cuts, repair captions, rebalance audio, or perform final creative polishes before rendering.