tldr

Download and summarize webpages using LLMs.

Installation

uv sync

Usage

tldr <URL> -m <provider/model>

Options

Option	Description
`-m, --model`	Model to use (required). Format: `provider/model`
`-p, --proxy`	Proxy URL (e.g., `http://proxy:8080`)
`-H, --header`	HTTP header (`Name: Value`). Repeatable
`-n, --nodriver`	Use headless browser for JS-rendered pages
`-d, --debug`	Show browser window when using `--nodriver`
`-r, --raw`	Output raw text without summarization
`-h, --help`	Show help

Models

tldr https://example.com -m anthropic/sonnet
tldr https://example.com -m openai/gpt-4o
tldr https://example.com -m gemini/flash
tldr https://example.com -m claude/sonnet
tldr https://example.com -m openrouter/...

Examples

# Basic usage
tldr https://example.com -m anthropic/sonnet

# JS-heavy pages
tldr https://example.com --nodriver -m claude/sonnet

# With proxy
tldr https://example.com --proxy http://localhost:8080 -m gemini/flash

# Custom headers
tldr https://example.com -H "Authorization: Bearer token" -m openai/gpt-4o

# Raw text output
tldr https://example.com --raw

# Debug mode (show browser window)
tldr https://example.com --nodriver --debug -m claude/sonnet

Environment Variables

Variable	Provider
`ANTHROPIC_API_KEY`	Anthropic
`OPENAI_API_KEY`	OpenAI
`GEMINI_API_KEY`	Google Gemini
`OPENROUTER_API_KEY`	OpenRouter

License

MIT