Zero-Cost Automated Social Media Pipeline

An open-source, zero-cost pipeline using n8n, Gemini AI, faster-whisper, and MoviePy to automatically generate captions, edit 9:16 clips, and post videos.

Comment

Inspiration

Creating short-form vertical content consistently is exhausting, but for creators with limited resources, it's even harder. Existing automation tools rely on expensive SaaS subscriptions or complex cloud setups like Docker. I wanted to build a 100% free, lightweight, privacy-focused, and self-hosted automation pipeline that handles the entire video processing and social media pipeline locally—without spending a single dollar.

What it does

The Zero-Cost Automated Social Media Pipeline turns a single raw video into a fully edited, captioned, and published short-form vertical video automatically.

  • One-Trigger Execution: A simple HTTP POST request or a tap inside Telegram kicks off the whole end-to-end process.
  • AI Copywriting & Subtitles: Uses faster-whisper for fast local speech transcription and Google Gemini to auto-write post copy.
  • Automated Video Editing: Transforms video into 9:16 vertical orientation via MoviePy & FFmpeg, cuts silent pauses, and burns time-aligned subtitles.
  • Multi-Platform Publishing: Guarantees video delivery via the Telegram Bot API and offers media staging for X (Twitter) using Playwright session reuse.

How I built it

  • Orchestration: Built on self-hosted n8n (Node.js) running locally on bare metal.
  • AI & NLP: Integrated Google Gemini API and faster-whisper for speech-to-text.
  • Video Rendering Pipeline: Written in Python 3.13 using MoviePy 2.2.1 and FFmpeg.
  • Bot & Publishing: Built a conversational Telegram bot (scripts/bot.py) and integrated Playwright for browser-automated X post staging.

Challenges I ran into

  • Hardware Constraints (The Biggest Struggle): My main challenge was my laptop. It simply wasn't powerful enough to run n8n, Python scripts, and Playwright browsers simultaneously. I suffered through constant system lags and freezes, which forced me to hyper-optimize memory usage (like using FFmpeg streaming filters) just to keep the pipeline alive.
  • The Hackathon Heartbreak: I managed to build this entire end-to-end system in a grueling 3-day sprint. Sadly, because my laptop struggled so much with rendering, I couldn't finish the demo video in time and completely missed the hackathon submission deadline right at the finish line.
  • Zero-Cost Engineering: Keeping the project completely free was a massive hurdle. Free-tier AI APIs kept hitting rate limits. I overcame this by engineering a custom key-rotation pool that seamlessly switches between multiple Gemini API keys whenever one gets blocked.

Accomplishments that I'm proud of

  • Built in 3 Days: Designing and coding a complete, functional orchestration pipeline from scratch in just 72 hours.
  • Overcoming Low-End Hardware: Successfully optimizing a heavy video-rendering and AI workflow to run on a very weak laptop.
  • Genuinely $0/Month: Achieving a truly free architecture by creatively bypassing API rate limits using multi-key rotation.

What I learned

  • How to optimize heavy tasks (like FFmpeg video rendering) for low-end hardware.
  • Time management under extreme pressure during a 3-day coding sprint.
  • How to creatively manage and bypass API rate limits using fallback rotation systems.

What's next for Social Media Automation

  • OpenClaw Integration (Zero API Costs Forever): Next, I plan to integrate OpenClaw—an open-source social media automation framework—to completely eliminate dependency on paid social media APIs. With OpenClaw handling the publishing layer, the system will seamlessly post to Instagram Reels, YouTube Shorts, TikTok, LinkedIn, and X without rate limits or subscription fees.
  • Long-Form Video Support: Expanding beyond short-form content to support full-length YouTube videos, complete with auto-generated chapter markers and custom thumbnail suggestions.
  • Cloud Compatibility: Creating an optimized version that can easily be deployed on a cheap cloud VPS for users who don't want to run it on their own hardware.

Built With

Share this project:

Updates

posted an update —

Project Status & Future Roadmap

Please note that this project is currently a work in progress and there is significant room for improvement.

Due to my current focus on my college/university admission journey, active development on my end is temporarily paused. I plan to return and implement major upgrades (including OpenClaw integration and long-form video support) after my admissions are completed.

Contributions are highly welcome! If you are a developer interested in making this zero-cost automation pipeline more robust, feel free to open an issue or submit a pull request. Let's build this better together!

Log in or sign up for Devpost to join the conversation.