FireRedTeam /FireRed-OpenStoryline

About FireRed-OpenStoryline

Talk to FireRed-OpenStoryline in plain language, and its LLM (a language model) plans the edit and runs the tools. You can search media, write narration, pick music and cut clips by chatting. API keys are required first, and AI transitions cost money.

Best for: Chat-first video makers who'd rather describe an edit than drag clips on a timeline.

• Searches and downloads images and clips, then splits and tags them
• Writes storylines and narration from a theme, matching a reference style
• Picks BGM with beat sync, plus voiceover tone and font suggestions
• Cuts, swaps and reorders clips by describing the change in chat
• Saves a whole edit as a Skill so you can reuse the style on new media
• Makes AI transitions from the last frame of one clip and first of the next

1. 1
   Clone the repo and enter it: the command below then cd FireRed-OpenStoryline

   git clone https://github.com/FireRedTeam/FireRed-OpenStoryline.git

   Copy
2. 2
   Make a Python 3.11 environment: the command below then conda activate storyline

   conda create -n storyline python=3.11

   Copy
3. 3
   Download models and install deps: sh build_env.sh on Linux/macOS, or ./download.sh then the command below

   pip install -r requirements.txt

   Copy
4. 4
   Add your API keys to config.toml (see docs/source/en/api-key.md)
5. 5
   Start the MCP server: the command below

   PYTHONPATH=src python -m open_storyline.mcp.server

   Copy
6. 6
   Open the chat UI: python cli.py for the CLI, or the command below for the web app

   uvicorn agent_fastapi:app --host 127.0.0.1 --port 8005

   Copy