🎬 YouTube Summarizer

Paste any YouTube URL — get an intelligent summary in seconds.

Supports any language · Full playlists · 4 summary styles · Timestamped key moments

(/frontend/src/assets/Screenshot02.png)

---

🧠 The Problem It Solves

People waste hours watching long YouTube videos just to extract a few key points. Creators post hour-long tutorials, lectures, and podcasts — but viewers often only need the core idea in 2 minutes.

YouTube Summarizer solves this by turning any YouTube video or playlist into a clean, structured summary — in seconds, in any language.

---

✨ Features

• 4 Summary Styles — Bullet points, short (3–5 sentences), detailed, or timestamped key moments
• Multilingual — Auto-detects spoken language (Urdu, Hindi, Spanish, etc.), always outputs English
• Playlist Support — Summarizes up to 10 videos in a playlist with a combined overview
• Smart Transcript Pipeline — Uses YouTube captions first (instant), falls back to Groq Whisper AI only when needed
• Long Video Support — Automatically chunks transcripts so even hour-long videos work
• In-Memory Cache — Same video won't be re-processed, responses are instant on repeat
• Download Summaries — Save any summary as a .txt file

---

🏗️ How It Works

YouTube URL
     │
     ▼
 FastAPI Backend (Render)
     │
     ├── Step 1: youtube-transcript-api ──► Captions found → instant transcript ✅
     │                                      (no download, no FFmpeg, works everywhere)
     │
     └── Step 2: Groq Whisper (fallback) ──► No captions? Download audio → transcribe
     │
     ▼
 Chunk long transcripts (prevents LLM context overflow)
     │
     ▼
 Groq LLaMA 3.3 70B (summarize with chosen style)
     │
     ▼
 React Frontend (Vercel) ──► Display + Download

Why not just use yt-dlp for everything? YouTube actively blocks audio downloads from cloud server IPs (Render, Railway, etc.). Using youtube-transcript-api as the primary method means the app reliably works in production. Whisper is kept as a real fallback for videos that genuinely have no captions.

---

🛠️ Tech Stack

Layer	Technology
Frontend	React 18 + Vite
Backend	FastAPI + Python 3.11
Transcription (primary)	youtube-transcript-api — no download needed
Transcription (fallback)	Groq Whisper whisper-large-v3
Summarization	Groq LLaMA 3.3 llama-3.3-70b-versatile
Backend hosting	Render.com (Docker)
Frontend hosting	Vercel

---

📁 Project Structure

youtube-summarizer/
├── backend/
│   ├── main.py                   # FastAPI app, routes, in-memory cache
│   ├── pipeline/
│   │   ├── transcript.py         # youtube-transcript-api (primary method)
│   │   ├── whisper_fallback.py   # Groq Whisper (fallback for no-caption videos)
│   │   ├── summarizer.py         # LLaMA 3.3 + transcript chunking logic
│   │   └── utils.py              # URL parsing, video ID extraction, playlist helpers
│   ├── requirements.txt
│   ├── Dockerfile
│   └── .env.example
├── frontend/
│   ├── src/
│   │   ├── App.jsx               # Main UI component
│   │   └── App.css               # Dark theme styling
│   ├── package.json
│   └── .env.example
├── render.yaml                   # Render deployment config
└── README.md

---

🚀 Run Locally

Prerequisites

• Python 3.11+
• Node.js 18+
• Groq API key (free)

1. Clone the repo

git clone https://github.com/muneeracodes/youtube-summarizer.git
cd youtube-summarizer

2. Backend

cd backend
python -m venv venv

# Windows
venv\Scripts\activate
# Mac/Linux
source venv/bin/activate

pip install -r requirements.txt
cp .env.example .env
# Edit .env → add your GROQ_API_KEY

uvicorn main:app --reload
# → http://localhost:8000

3. Frontend

cd frontend
npm install
cp .env.example .env.local
# Edit .env.local → set VITE_API_URL=http://localhost:8000

npm run dev
# → http://localhost:5173

---

☁️ Deploy Your Own

Backend → Render.com

1. Go to render.com → New → Web Service
2. Connect your GitHub repo
3. Set Root Directory → backend
4. Set Environment → Docker
5. Add env var: GROQ_API_KEY = your key
6. Add env var: ALLOWED_ORIGINS = your Vercel URL
7. Deploy → copy the service URL

Frontend → Vercel

1. Go to vercel.com → Add New Project → import repo
2. Set Root Directory → frontend
3. Add env var: VITE_API_URL = your Render URL
4. Deploy

---

🔑 Environment Variables

Backend (backend/.env)

Variable	Required	Description
GROQ_API_KEY	✅	Free at console.groq.com
ALLOWED_ORIGINS	✅	Your Vercel URL (for CORS)

Frontend (frontend/.env.local)

Variable	Required	Description
VITE_API_URL	✅	Your Render backend URL

---

📬 API Reference

POST /summarize

{
  "url": "https://www.youtube.com/watch?v=VIDEO_ID",
  "style": "bullets"
}

Styles: bullets · short · detailed · timestamps

Response:

{
  "video_id": "VIDEO_ID",
  "title": "Video Title",
  "summary": "- Key point one\n- Key point two...",
  "transcript_method": "captions",
  "cached": false
}

POST /summarize/playlist

{
  "url": "https://www.youtube.com/playlist?list=PLAYLIST_ID",
  "style": "short"
}

GET /health

Returns { "status": "ok" } — used by Render for health checks.

---

👩‍💻 Author

Muneera Ibrahim