The short drama market is exploding, but creators hit the same wall: slow production, high costs, and privacy leaks. Traditional routes mean hiring writers and editors (days of work) or paying $30-70/month for cloud AI tools that can expose your commercial secrets. LocalMiniDrama changes the game. This open-source, local-first AI short drama tool runs a full pipeline from story generation to video synthesis. Plug in your own AI API keys — Tongyi, Volcano, Kling, Gemini — and produce professional short dramas in hours, not weeks. The GitHub repo shows 1,736 stars and 421 forks under an MIT license. Community momentum is real.

This tool demolishes the entry barrier. Most AI drama platforms rely on cloud services. LocalMiniDrama is built from scratch in pure JavaScript. Bring your own API keys, zero subscription fees. It even supports a canvas mode — manage your storyboard pipeline like building blocks. The core selling point: local-first. SQLite storage, assets never touch the cloud, fully open source for customization. No third-party data theft. No cloud API bills. Compare that to tools charging hundreds per month with less flexibility. LocalMiniDrama gets you producing professional short dramas at near-zero cost — for client work, content operations, or selling digital products.

Core Advantage: Local-First, Privacy Under Your Control

The biggest win: everything runs locally. Unlike most AI short drama tools, all generated results stay on your machine. No cloud dependency. That kills the privacy problem — critical for commercial projects or personal storytelling. The project uses SQLite plus local file storage. Assets never leave your machine. You own your work completely. The developer organization is xuanyustudio, repo at https://github.com/xuanyustudio/LocalMiniDrama. Official description: “Data never leaves your machine. Short drama workflow management platform. High flexibility. AI live-action and AI comic dramas handled locally.”

One caveat: generated results stay local, but model calls send data to API providers (Tongyi, Volcano, etc.). Assess privacy risks yourself. The official docs mention this too.

Beyond local-first, the pipeline is fully automated. Eight core functions: story generation, script editing, character generation, scene generation, prop generation, storyboard generation, image/video generation, and video synthesis. One-click pipeline mode intelligently skips completed content. Failed steps auto-retry three times. Real-time progress and error logs. The latest v1.2.8 adds Agnes AI integration, enhanced canvas mode, ModelArk private asset library config, and configurable image hosting. Prompt optimization and storyboard image count fixes make the tool increasingly mature.

Multi-model support deserves attention. Text, image, and video models configure independently. One-click setup for Tongyi and Volcano. Nine prompt categories are fully customizable. Mix and match models to boost output quality. Use Volcano Seedance 2.0 for video, Tongyi Wanxiang for character images, Kling Omni for complex action. Compared to single-model tools, this flexibility gets you closer to live-action quality with consistent character appearances. The built-in wuxia short drama example shows consistent costumes and characters across storyboards — proof the tool delivers.

The community side: canvas workflow docs, drag-and-drop coordinate saving, curve connections, box-select batch operations. Layout persistence keeps your storyboard edits safe. Right-click menus and floating toolbars speed up workflow. LocalMiniDrama’s edge: open source + local + multi-model + full automation. Together, they form an offline creation loop that barely exists elsewhere. Lower entry barrier, infinite expansion space.

Installation & Setup: Zero Experience Welcome

Download the latest exe from https://github.com/xuanyustudio/LocalMiniDrama/releases. Pick the standard exe with sample projects — it’s a reasonable size and includes preset examples. Perfect for beginners. Double-click to run. First launch shows the AI configuration screen.

Fill in your API keys in config.yaml. That’s the prerequisite. Recommended models:

• Tongyi Qianwen or Volcano Engine for text and image generation
• Kling or Volcano Seedance 2.0 for video clips
• ModelArk private asset library with Volcano Ark AK/SK dual authentication

Setup is simple: edit configs/config.yaml in the backend directory. Write upload_url, timeout (default 180 seconds), and retry count. Save and exit. The whole process takes under 15 minutes. Official quick-start guide and AI configuration docs are available. After first run, you’ll see the project card list on the local dashboard.

Note: this tool requires some technical basics (like editing config files). If you have zero tech background, learn basic config file editing first, or follow official docs and community tutorials. Detailed dev/packaging/Docker guides help you ramp up.

Want to develop from source? Repo at https://github.com/xuanyustudio/LocalMiniDrama. Node.js 18+. Clone, cd