Full-stack YouTube research platform for creators and content teams. YT Analyzer helps find outlier videos, analyze transcripts, compare trends, research keywords, and generate AI-assisted content ideas.
Built with Next.js 16, React 19, TypeScript, Tailwind CSS 4, Prisma, SQLite, Google Gemini, YouTube Data API, and Google Trends.
- Outlier Finder - Search YouTube by keyword, filter by views/subscribers/duration/engagement, and surface unusually strong videos.
- Video Analyzer - Paste YouTube URLs, fetch transcripts and metadata, and generate deep analysis of hooks, structure, thumbnails, and content patterns.
- Keyword Research - Combine YouTube autocomplete with AI keyword generation and brainstorming.
- Google Trends - Compare keyword interest over time and inspect rising queries and regional interest.
- AI Tools - Generate ideas from saved folders and summarize videos.
- Folders - Save videos into collections for later research and batch analysis.
Requires a Google Cloud project with YouTube Data API v3 enabled. Set the project ID and region through environment variables.
| API | Used for | Auth |
|---|---|---|
| YouTube Data API v3 | Video search, metadata, and channel stats | YOUTUBE_API_KEY |
Gemini via @google/genai |
Analysis, summaries, keyword generation, and idea generation | Google project/location environment |
The app also uses public/non-key integrations:
- Google Trends via
google-trends-api - YouTube autocomplete via the public suggest endpoint
- YouTube transcript extraction via
youtube-transcript
Create a local .env file:
DATABASE_URL="file:./dev.db"
YOUTUBE_API_KEY=your_youtube_data_api_v3_key
GOOGLE_PROJECT_ID=your_google_cloud_project_id
GOOGLE_CLOUD_LOCATION=us-central1npm install
npx prisma migrate dev
npm run devOpen http://localhost:3000 to use the app locally.