67

  ████████   ██████████
 ███░░░░███ ░███░░░░███
░███   ░░░  ░░░    ███ 
░█████████        ███  
░███░░░░███      ███   
░███   ░███     ███    
░░████████     ███     
 ░░░░░░░░     ░░░      

A fast, focused YouTube downloader CLI written in Rust.

Features

• 🚀 Fast - Direct InnerTube API, no browser overhead
• 📦 Batch downloads - Concurrent downloads from URL lists
• 📝 Transcript extraction - Extract subtitles/captions without downloading video
• 🎨 Cinematic banner - Rainbow gradient ASCII art
• 📊 Format selection - Best, worst, audio-only, video-only, or specific format IDs
• ⏸️ Resume support - Interrupted downloads continue via .part files
• 🔇 Quiet mode - For scripting and automation

Quick Start

# Download a video
67 "https://www.youtube.com/watch?v=dQw4w9WgXcQ"

# List available formats
67 -F VIDEO_ID

# Download audio only
67 --audio-only VIDEO_ID

# Batch download with 4 concurrent jobs
67 -b urls.txt -j 4 -O ./downloads

Installation

git clone https://github.com/YOUR_USERNAME/67.git
cd 67
cargo build --release

# Binary at target/release/67

Usage

# Download best quality
67 VIDEO_ID
67 "https://www.youtube.com/watch?v=VIDEO_ID"
67 "https://youtu.be/VIDEO_ID"

# List formats
67 -F VIDEO_ID

# Select specific format
67 -f 22 VIDEO_ID
67 -f bestaudio VIDEO_ID
67 -f bestvideo VIDEO_ID

# Audio only
67 --audio-only VIDEO_ID

# Custom output
67 -o "my_video.mp4" VIDEO_ID

# Batch download
67 -b urls.txt -j 4 -O ./downloads

# Quiet mode (for scripts)
67 -q VIDEO_ID

# Extract transcripts (no download)
67 -M VIDEO_ID

Transcript Extraction

Extract subtitles/captions to JSON without downloading the video:

67 -M VIDEO_ID
67 -M "https://www.youtube.com/watch?v=VIDEO_ID"
67 -M VIDEO_ID -o transcripts.json

Output format:

{
  "video_id": "dQw4w9WgXcQ",
  "title": "Video Title",
  "channel": "Channel Name",
  "transcripts": [
    {
      "language": "English",
      "language_code": "en",
      "is_auto_generated": false,
      "segments": [
        {"start": 1.36, "duration": 1.68, "text": "Hello world"}
      ]
    }
  ]
}

Format Selection

Selector	Description
best	Best muxed format (video+audio)
bestaudio	Best audio-only format
bestvideo	Best video-only format
worst	Lowest quality muxed format
137	Specific itag (from -F listing)

Batch Downloads

Create a file with URLs (one per line):

# urls.txt
https://youtube.com/watch?v=VIDEO1
https://youtube.com/watch?v=VIDEO2
# Comments are ignored
VIDEO3

Download with concurrency:

67 -b urls.txt -j 4 -O ./downloads

How It Works

URL → Video ID → InnerTube API → Parse Formats → Concurrent Download

1. URL Parsing - Extracts video ID from any YouTube URL format
2. InnerTube API - Direct API call (Android client) bypasses throttling
3. Format Selection - Parses streamingData.formats and adaptiveFormats
4. Download - Chunked HTTP with resume support and progress bar

Project Structure

src/
├── main.rs        # CLI entry point & batch processing
├── banner.rs      # Cinematic ASCII banner
├── extractor.rs   # InnerTube API extraction  
├── formats.rs     # Format parsing and selection
├── downloader.rs  # HTTP download with progress
├── metadata.rs    # Transcript extraction
└── error.rs       # Error types

Dependencies

• reqwest - Async HTTP client with connection pooling
• tokio - Async runtime
• futures - Concurrent stream processing
• tui-banner - Cinematic terminal banners
• clap - CLI argument parsing
• indicatif - Progress bars
• console - Terminal styling

Why "67"?

Short. Memorable. Slightly mysterious. 🎲

License

MIT