lightpanda-io/browser

Lightpanda: the headless browser designed for AI and automation

View on GitHub ↗Jump to charts ↓

Summary Information

Updated 12 minutes ago
Added to GitGenius on March 22nd, 2026
Created on February 7th, 2023
Open Issues & Pull Requests: 94 (+0)
GitHub issues: Enabled
Number of forks: 1,675
Total Stargazers: 35,323 (+0)
Total Subscribers: 95 (+0)

Repository Insights (GitGenius)

Median issue/PR response: 16.8 hours
Mean response time: 45.9 days
90th percentile: 75.3 days
Tracked items: 478

Most active contributors

Sign in to see contributor activity.

How this project is maintained

Around half of the issues opened in the past year never receive a reply. 89% of open issues come from outside the core team, so the backlog reflects real-world use rather than internal planning. Work labelled "bug" is answered fastest, typically in about 9 hours, while "DOM API" waits about 9 months. 59% of tracked open issues have had no activity in three months. Only 12% of issues opened in the past year have been closed.

Charts & Analytics

Fetching additional details & charts...

Issue Activity (beta)

Issue API getrepoissuespagesummary failed: 429 Rate limit exceeded. Please try again later.

Detailed Description

Lightpanda is a headless browser written entirely in Zig, designed specifically for AI agents and automation tasks. Unlike existing solutions, it is not a fork of Chromium or a patch to WebKit, but rather a completely new browser implementation built from the ground up. The project is currently in Beta status and actively maintained, with the primary goal of providing a lightweight, fast alternative to traditional headless browsers for server-side automation and web scraping.

The repository demonstrates significant performance advantages over headless Chrome in benchmark tests conducted on real web pages. When processing 100 pages, Lightpanda uses approximately 123MB of peak memory compared to Chrome's 2GB, representing roughly 16 times less memory consumption. Execution time shows even more dramatic improvements, completing the same task in 5 seconds versus Chrome's 46 seconds, approximately 9 times faster. These benchmarks were generated by requesting 933 real web pages over the network on an AWS EC2 m5.large instance, providing realistic performance metrics for production scenarios.

The browser supports multiple installation methods including Homebrew, the Arch Linux User Repository, and nightly binary builds for Linux and macOS on both x86_64 and aarch64 architectures. Official Docker images are available for both Linux amd64 and arm64. The project provides a Chrome DevTools Protocol server that can be accessed on port 9222, enabling compatibility with existing automation tools like Puppeteer and Playwright. Additionally, Lightpanda includes an agent mode that allows users to control the browser through natural language instructions, supporting multiple LLM providers including Anthropic, OpenAI, Gemini, Google Vertex AI, Hugging Face, and local models via Ollama.

Core technical features include HTTP loading via Libcurl, HTML parsing through html5ever, a complete DOM tree implementation, JavaScript execution powered by V8, and comprehensive DOM API support. The browser handles Ajax requests, XHR API, Fetch API, DOM dumping, WebSocket connections, form input, cookies, custom HTTP headers, proxy support, and network interception. It also respects robots.txt when configured with the appropriate option. The project uses PandaScript, a vanilla JavaScript variant with native browser primitives, allowing users to export and replay automation scripts deterministically without requiring runtime LLM access.

The project requires Zig version 0.15.2 for building from source and depends on V8, Libcurl, and html5ever. Build infrastructure includes unit tests, end-to-end tests using a separate demo repository, and Web Platform Tests for standards compliance. The project includes telemetry collection by default, which can be disabled via the LIGHTPANDA_DISABLE_TELEMETRY environment variable. Development is guided by a Contributor License Agreement, and the community maintains an active Discord server for collaboration and support.