Building a YouTube Clickbait Checker Web App with Python and AI - Source Code

This source code allows you to build a full-featured YouTube Clickbait Checker using AI technologies.

The web application downloads YouTube videos, transcribes the audio, and analyzes the video title against its transcript to determine the likelihood of clickbait.

Powered by Deepgram for speech-to-text transcription and Mistral AI for text analysis, this app provides an intuitive way to evaluate YouTube titles and display a clickbait score with reasoning.

Key Features:

• AI-Powered Title Analysis: Leverages Mistral AI to compare video titles with transcripts and returns a clickbait score.

• Automatic Transcription: Uses Deepgram’s API to automatically transcribe video audio.

• YouTube Video Processing: Downloads YouTube videos and extracts audio using yt-dlp.

• Web Interface: Clean, responsive frontend where users can input a YouTube URL and receive real-time results.

• Error Handling: Robust error-handling for failed transcriptions or API requests.

Technologies Used:

• Flask (Python) for backend API.

• yt-dlp for video downloading and audio extraction.

• Deepgram API for speech-to-text conversion.

• Mistral AI for title-to-transcript comparison.

• HTML/CSS/JavaScript for a simple user interface.

Ideal for developers looking to integrate AI-based content analysis into their projects, this code provides a solid foundation for building or expanding clickbait detection tools.