glance-cli

glance-cli

AI-powered web reader for your terminal – Fetch any webpage, extract clean content, get instant AI summaries, and listen with natural voice synthesis.

Why glance-cli?

• ✅ 100% FREE – Uses local Ollama (no API keys needed)
• 🌍 Auto language detection – English, French, Spanish, Haitian Creole
• 🔒 Privacy first – Your browsing history stays local
• 🎤 Voice-enabled – Read articles aloud with multilingual support
• 💾 Multiple formats – Save as markdown, JSON, or plain text
• ⚡ Lightning fast – Built with Bun and TypeScript

Quick Start

# Install
npm install -g glance-cli

# Use immediately
glance https://news.com                    # Get AI summary
glance https://news.com --read             # Read aloud
glance https://news.com --browse           # Interactive navigation

For 100% free local AI:

curl -fsSL https://ollama.com/install.sh | sh
ollama pull llama3
glance https://techcrunch.com --tldr --read

Core Features

🌐 Interactive Browse Mode

Navigate websites interactively with AI summaries on demand.

glance https://news.ycombinator.com --browse

# Inside browse mode:
1                    # Navigate to link 1
3 --tldr            # Navigate + get summary
5 --read            # Navigate + read aloud
1 --output file.md  # Navigate + save to file
n                   # Show navigation links
e                   # Show external links
b                   # Go back
q                   # Quit

📝 Read & Summarize

Extract content with AI summaries in multiple styles.

glance <url>              # AI summary (default)
glance <url> --tldr       # One sentence
glance <url> --key-points # Bullet points
glance <url> --eli5       # Simple explanation
glance <url> --full       # Full content (no summary)
glance <url> --copy       # Copy to clipboard

🎤 Voice & Audio

Listen to articles with natural voice synthesis.

glance <url> --read                    # Read aloud (auto-detects language)
glance <url> --audio-output audio.mp3  # Save as MP3
glance <url> --voice nova --read       # Choose specific voice
glance <url> -l fr --read              # French voice

glance --list-voices                   # See available voices

Voice Options:

• Free: System TTS (macOS/Windows/Linux)
• Premium: ElevenLabs (natural voices, requires API key)

💾 Save & Export

Save content in multiple formats.

glance <url> --output summary.md       # Save as markdown
glance <url> --output data.json        # Save as JSON
glance <url> --format plain -o file    # Save as plain text

🌍 Auto Language Detection

Automatically detects and adapts to content language.

glance https://lemonde.fr --tldr       # Auto-detects French
glance https://news.es --read          # Auto-detects Spanish + voice
glance <url> -l es --read              # Override to Spanish

Supported: English, French, Spanish, Haitian Creole

Common Use Cases

# Morning news with audio
glance https://news.ycombinator.com --tldr --read

# Quick copy for sharing
glance https://article.com --tldr --copy

# Interactive research
glance https://en.wikipedia.org/wiki/AI --browse

# Study while coding
glance https://tutorial.com --full --read

# Save documentation
glance https://docs.python.org --output guide.md

# Multilingual content
glance https://lemonde.fr --read        # Auto-detects French voice

AI Models

Local (Free) – via Ollama:

• llama3:latest – Fast, reliable (recommended)
• gemma3:4b – Lightweight, efficient
• mistral:7b – Excellent quality
• deepseek-r1:latest – Strong reasoning

Cloud (Optional) – requires API keys:

• gpt-4o-mini – OpenAI
• gemini-2.0-flash-exp – Google Gemini

glance <url> --model llama3            # Use specific model
glance --list-models                   # Show available models
glance <url> --free-only               # Never use paid APIs

Configuration

# AI Providers (optional)
export OPENAI_API_KEY=...
export GEMINI_API_KEY=...

# Voice (optional, for premium voices)
export ELEVENLABS_API_KEY=...

# Ollama (optional, auto-detected)
export OLLAMA_ENDPOINT=http://localhost:11434

All Options

# Content
--browse                 # Interactive navigation
--tldr                   # One sentence summary
--key-points, -k         # Bullet points
--eli5                   # Simple explanation
--full                   # Full content
--ask "question"         # Custom Q&A

# Voice & Audio
--read, -r               # Read aloud
--audio-output <file>    # Save as MP3
--voice <name>           # Choose voice
--list-voices            # Show available voices
-l, --language <code>    # Output language (en, fr, es, ht)

# AI Models
--model <name>           # Choose AI model
--list-models            # Show local models
--free-only              # Never use paid APIs

# Output
--format <type>          # Format: md, json, plain
--output, -o <file>      # Save to file
--copy, -c               # Copy to clipboard
--stream                 # Live streaming output

# Advanced
--screenshot <file>      # Capture screenshot
--full-render            # Render JavaScript
--metadata               # Show page metadata
--links                  # Extract all links
-v, --verbose            # Detailed logs

Contributing

Contributions welcome! Check out our Contributing Guide.

git clone https://github.com/jkenley/glance-cli.git
cd glance-cli
bun install
bun dev https://example.com

Support

If you find glance-cli useful, consider buying me a coffee ☕

Links

• GitHub: github.com/jkenley/glance-cli
• NPM: npmjs.com/package/glance-cli
• Issues: github.com/jkenley/glance-cli/issues

Built with ❤️ by Kenley Jean • MIT License

Star ⭐ the repo if you find it useful!