Skip to content

End-to-end pipeline to scrape clips, auto-describe with GPT, generate overlays, and render thousands of short videos hands-free.

Notifications You must be signed in to change notification settings

Appilot123/GPT-Short-Video-Mass-Generator

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 

Repository files navigation

GPT Short Video Mass Generator

Project Overview:

This repo wraps a full pipeline for pumping out thousands of short-form videos—end to end. It scrapes clips, grabs screenshots, sends frames to GPT for smart descriptions, auto-generates multiple on-brand text overlays, then burns everything back onto the videos with FFmpeg/Python. Perfect for bulk content farms for TikTok/Reels/Shorts without babysitting every edit.

Core Features:

  • Hands-off throughput: One command spins the whole chain—scrape → analyze → caption → render.
  • Human-ish creativity at scale: GPT handles semantic tagging and overlay copy so videos don’t feel templated.
  • Stealthy, resilient scraping: Rotating proxies + randomized timing to keep sources happy and IPs fresh
select device
Have a Custom Project in Mind please Contact?
Email support@appilot.app    Book a 30-minute Call Book a 30-minute Call

Features List:

Feature Description
Video Scraper Pulls thousands of niche-relevant clips from defined sources with proxy rotation.
Auto Screenshotter Extracts 3 representative frames per video for semantic analysis.
GPT Frame Describer Feeds frames to OpenAI API to generate context-aware summaries.
Overlay Text Generator Produces 5 punchy overlay lines per clip, tuned to your niche.
FFmpeg Batch Renderer Stacks overlays onto videos in bulk with position/size randomization.
Retry & Queue System Fault-tolerant jobs with resumable queues for long runs.
Config-First Design YAML/ENV driven: paths, API keys, model choices, overlay styles.
Proxy & Rate Control Rotates proxies, inserts jittered delays to mimic human traffic.
Audit Logs & Reports CSV/JSON logs of each step (scraped, processed, failed, rendered).
Modular Pipeline Swap GPT provider, scraper, or renderer without gutting the code.

Key Stats:

  • Automation Accuracy: 98% successful end-to-end runs
  • Session Stability: 24/7 processing with auto-resume
  • Concurrent Account/Sources: 200+ handled via queue shards
  • Action Randomization Effectiveness: 99% pass rate in internal A/B tests

Tech & Stack:

  • Python (scraping, orchestration), FFmpeg (video ops)
  • OpenAI/LLM APIs for descriptions & overlays
  • Playwright/Selenium + Rotating Proxies for stealth grabs
  • Optional: Redis/RQ for job queues, Docker for reproducible runs

About

End-to-end pipeline to scrape clips, auto-describe with GPT, generate overlays, and render thousands of short videos hands-free.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published