v0.1.0 - First working version
This is the first public release of ScriptSmith-Engine, an AI video pipeline that turns a text script into a narrated YouTube-style video.
Features
- Generate narration using Gemini TTS
- Create transcription timestamps with faster-whisper
- Plan scenes using MiniMax
- Generate AI images using Runware / GPT Image
- Search and add sound effects from Freesound
- Assemble the final video using FFmpeg
Notes
This release is still experimental. You will need to configure your own API keys in .env.
Setup
- Clone the repo
- Install dependencies from
requirements.txt - Copy
.env.exampleto.env - Add your API keys
- Run the pipeline scripts