Skip to content

README.md

Latest commit

 

History

History
138 lines (102 loc) · 3.91 KB

README.md

File metadata and controls

138 lines (102 loc) · 3.91 KB
Raw

CookVid

AI-powered cooking video editor. Automatically cuts, stitches, and cleans up cooking videos using Claude AI — no manual editing required.

Features

  • Cut to Short — Takes a long-form cooking video and creates a 30-60 second short by identifying the most engaging segment
  • Stitch Videos — Combines multiple cooking videos into a cohesive long-form video or short
  • Edit Long-form — Cleans up a long-form video by removing dead air, filler words, bad takes, and off-topic tangents

How It Works

Video → FFmpeg extracts audio → Whisper transcribes → Claude analyzes transcript → FFmpeg applies edits → Output
  1. Audio is extracted from the video using FFmpeg
  2. Whisper transcribes the audio locally with timestamps
  3. The transcript is sent to Claude, which returns a structured edit plan (what to keep, cut, and reorder)
  4. MoviePy/FFmpeg executes the edits and produces the final video

Prerequisites

Install FFmpeg

# macOS
brew install ffmpeg

# Ubuntu/Debian
sudo apt install ffmpeg

# Windows
winget install ffmpeg

Setup

# Clone the repo
git clone <repo-url>
cd video-editor-tool

# Create and activate virtual environment
python3 -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

# Configure your API key
cp .env.example .env
# Edit .env and add your ANTHROPIC_API_KEY

Usage

CLI

# Cut a long video into a short
python -m app.cli cut-to-short cooking_video.mp4 -o short.mp4

# Stitch multiple videos into a long-form video
python -m app.cli stitch video1.mp4 video2.mp4 video3.mp4 -o combined.mp4

# Stitch multiple videos into a short
python -m app.cli stitch video1.mp4 video2.mp4 --type short -o highlight.mp4

# Clean up a long-form video
python -m app.cli edit-longform raw_footage.mp4 -o cleaned.mp4

# Use a different Whisper model for better accuracy
python -m app.cli cut-to-short video.mp4 --whisper-model medium

Web UI

python -m app.cli serve

Open http://127.0.0.1:8000 in your browser. Upload videos, choose an editing mode, and download the result.

Options

Option Description Default
-o, --output Output file path ./output/<name>_<type>.mp4
--whisper-model Whisper model size: tiny, base, small, medium, large base
--type Stitch output type: short or longform longform
--host Web server host 127.0.0.1
--port Web server port 8000

Configuration

Environment variables (set in .env):

Variable Description Default
ANTHROPIC_API_KEY Your Anthropic API key required
WHISPER_MODEL Default Whisper model size base
OUTPUT_DIR Directory for output files ./output

Project Structure

video-editor-tool/
├── config.py              # Settings via environment variables
├── app/
│   ├── cli.py             # Click CLI commands
│   ├── web.py             # FastAPI web interface
│   ├── models/
│   │   └── schemas.py     # Pydantic data models
│   ├── pipeline/
│   │   ├── audio.py       # FFmpeg audio extraction
│   │   ├── transcribe.py  # Whisper transcription
│   │   ├── analyze.py     # Claude API integration
│   │   ├── edit.py        # MoviePy video editing
│   │   └── runner.py      # Pipeline orchestrator
│   ├── prompts/
│   │   ├── cut_to_short.py
│   │   ├── stitch.py
│   │   └── edit_longform.py
│   └── templates/
│       └── index.html     # Web UI
├── requirements.txt
├── setup.py
└── .env.example

License

MIT