An end-to-end autonomous pipeline that transforms Reddit stories into professionally edited short-form videos with zero manual intervention.
Creating short-form content manually takes hours. Searching for stories, recording voiceovers, syncing subtitles word-by-word, and rendering videos is a repetitive nightmare.
ViralContent Factory connects 5 AI services, TTS engines, and a programmable video compositor into one seamless workflow.
Manages execution, handles batching, and prevents system sleep.
Scraping & AI LLM processing for hooks and SEO. Checks 30+ subreddits.
Audio synthesis via Edge-TTS & word-level timestamp extraction.
MoviePy video compositing with dynamic 3-word animated subtitles.
Automatic failover between Groq, Gemini, Cerebras, HF, and OpenRouter.
Batch management, removing temporary files, email alerts when 7+ ready.
Automated YouTube and Instagram modules.
Never hit a rate limit. Automatically falls back between 5 LLM providers if quotas are exhausted.
Mathematically accurate word-level alignment with configurable offset mapping.
Scrapes top engaging stories seamlessly.
Runs quietly in the background and sends an email when 7 videos are ready.
Built-in OS sleep prevention ensures long renders finish.
No storage bloat. Temp files are deleted instantly.
The built-in LLM router prevents your pipeline from crashing due to API limits. It automatically uses cheap fast models for basic tasks and strong models for creative hooks.
You need FFmpeg, ImageMagick, and Deno (for yt-dlp).
Task Scheduler triggers the factory. Scraper randomly selects from 30+ subreddits, filters out bad posts, and finds the perfect 150-word story.
LLM Router connects to Groq. Generates a viral hook, detects gender, assigns voice, and creates 5 SEO tags. A/B testing ranks the best title.
Edge TTS generates high-quality neural narration. System extracts mathematically precise word-level timestamps with a -0.3s sync offset.
MoviePy composites random 4K gameplay footage in 9:16 vertical format. Dynamic 3-word animated subtitles pop on screen using ImageMagick.
3 videos are completed. The system deletes temporary audio/json files. Checks if 7 videos are ready. Sends a Gmail alert.
Star the repository, fork the code, and start generating your first batch of viral videos today.
Star on GitHub