Problem / Context
Turning researched text into cinema-quality 16:9 narrative videos with synchronized voiceovers.
What it does
- Text → scenes (LLM or rule-based), cinematic prompt enhancement
- Multi-language TTS (edge-tts/Kokoro), pluggable video gen (Mock/Wan/Hunyuan/LTX)
- FFmpeg assembly, word-by-word subtitles, checkpoint/resume, dry run, publishing hook
Engineering notes
Python 3.10+ CLI pipeline with provider plugins; Docker option for local video gen; zero-cost default stack.
Outcome / Value
End-to-end AI video generation pipeline for content-automation clients.
Screens
