Skip to content

OpenDirector v0.1.0

Choose a tag to compare

View all tags
@hsliuustc0106 hsliuustc0106 released this 06 Nov 06:23
v0.1.0
76f61e7

OpenDirector v0.1.0

The first versioned release of OpenDirector, a local, browser-based AI video creation workspace. Generate assets through chat, organize references on a canvas, edit media, and arrange a timeline for export.

Highlights

  • AI-assisted creation with configurable model services, reference assets, and generation history.
  • A canvas for uploaded and generated assets, with connections and @ references.
  • Image, video, and audio editing, plus image presets for character, face and product reference sheets, storyboards, and multi-camera variations.
  • Timeline editing with trimming, speed and volume controls; MP4, JSON, FCPXML, and asset-bundle export.
  • Local project storage, a business CLI, and Chinese and English documentation.
  • Source installation through python -m pip install -e ., including the frontend build, followed by opendirector serve.

Included refinements

  • Expanded image presets and a simpler generation panel (#124).
  • Separate smart-edit marks and selections, including nonempty-mask validation (#125).
  • Restored project canvas links and improved project-history reads (#126).
  • Direction indicators on canvas connection ports (#127).
  • Bounded local video clip recording and stalled-playback detection (#128).

Getting started

Source installation requires Python 3.11+, Node.js 24.22.1+ with npm, Git, and FFmpeg/ffprobe with libx264. Configure your chosen model services in Settings → Model connections, or upload local media to begin editing.

See the installation guide and user guide.

Limitations and validation

  • MP4 export requires Chrome or Edge with the page and editor kept open. This release does not claim browser-independent background rendering.
  • AI features depend on the capabilities, availability, credentials, and terms of the services you configure. Local project storage does not mean generation requests remain on your machine.
  • No GitHub Actions runs or check runs were recorded for the release commit when prepared. The release preparation did not execute local tests, real-provider generation, or an end-to-end export; it is not a certification that all workflows pass.
  • Model-connection credential handling when changing a service URL needs further validation. Use credentials intended for the configured endpoint.
  • The example-film README credits third-party music. Redistribution rights for that media were not verified in this release preparation; confirm them before wider redistribution.
  • Code licensing is described in LICENSE; third-party notices are in NOTICE.

Release source: 76f61e769aa943dc79a71ac9604b3893b522657d, from release/v0.1.0.

中文说明

OpenDirector 的首个版本发布：本地运行、通过浏览器使用的 AI 视频创作工作区。通过对话生成素材，在画布上组织引用，编辑媒体，再通过时间线编排与导出。

主要功能

  • 配置模型服务，通过 AI 助手创作，使用参考素材并查看生成历史。
  • 在画布上管理上传与生成的素材，通过连线和 @ 引用建立参考关系。
  • 图片、视频和音频编辑，以及人物、面部、产品三视图、分镜和多机位图片预设。
  • 时间线剪辑、速度和音量调整，支持 MP4、JSON、FCPXML 和素材包导出。
  • 本地工程存储、业务 CLI，以及中英文文档。
  • 通过 python -m pip install -e . 安装并自动构建前端，使用 opendirector serve 启动。

本版本包含图片预设与生成面板优化（#124）、智能编辑标记和选区隔离（#125）、工程链接与历史读取修复（#126）、画布连接方向提示（#127），以及本地视频片段录制时长与播放停滞处理（#128）。

安装与使用边界

源码安装需要 Python 3.11+、Node.js 24.22.1+ 与 npm、Git，以及支持 libx264 的 FFmpeg/ffprobe。模型服务需自行配置，也可以先上传本地素材开始编辑。

MP4 导出请使用 Chrome 或 Edge，并保持页面和编辑器开启；本版本不承诺脱离浏览器的后台渲染。AI 能力取决于所配置的服务，生成请求可能发送至外部服务。

准备发布时，该提交没有 GitHub Actions 或 check run 记录；本次准备未执行本地测试、真实模型生成或完整导出验收，不代表所有流程已验证通过。更改模型服务地址时的凭证处理仍需进一步验证，请使用与目标地址对应的凭证。示例影片包含注明的第三方音乐，本次未核验其再分发授权，扩大传播前需另行确认。代码许可与第三方声明请参阅 LICENSE 和 NOTICE。

Assets 2
Loading