Clone Viral TikTok & Instagram Reels with Apify and Gemini 2.5 Pro

Reverse engineer short-form videos from Instagram and TikTok using Gemini AI

Who's it for

Content creators, AI video enthusiasts, and digital marketers who want to analyze successful short-form videos and understand their production techniques. Perfect for anyone looking to reverse-engineer viral content or create detailed prompts for AI video generation tools like Google Veo or Sora.

How it works

This automation takes any Instagram Reel or TikTok URL and performs a forensic analysis of the video content. The workflow downloads the video, converts it to base64, and uses Google's Gemini 2.5 Pro vision API to generate an extremely detailed "Generative Manifest" - a comprehensive prompt that could be used to recreate the video with AI tools.

The analysis includes: Visual medium identification (film stock, camera sensor, lens characteristics) Color grading and lighting breakdown Shot-by-shot deconstruction with precise timing Camera movement and framing details Subject description and action choreography Environmental and atmospheric details

How to set up

Configure API credentials: Add your Apify API key for video scraping Set up Google Gemini API authentication

Set up Slack integration (optional): Configure Slack OAuth for result sharing Update the channel ID where results should be posted

Access the form: The workflow creates a web form where you can input video URLs Form accepts both Instagram Reel and TikTok URLs

Requirements

Apify account** with API access for video scraping Google Cloud account** with Gemini API enabled Slack workspace** (optional, for sharing results) Videos must be publicly accessible (no private accounts)

How to customize the workflow

Modify the analysis prompt:** Edit the "set_base_prompt" node to adjust the depth and focus of the video analysis Add different platforms:** Extend the switch node to handle other video platforms Integrate with other tools:** Replace Slack with email, Discord, or other notification systems