- Goose, a free open-source AI coding agent from Block, has amassed over 42,000 GitHub stars as a Claude Code alternative.
- The tool works with 15+ LLM providers and 70+ MCP extensions—Claude Code locks you into Anthropic’s ecosystem.
- Goose just moved to the Linux Foundation’s Agentic AI Foundation, signaling it’s not going away anytime soon.
Claude Code is the most talked-about AI coding agent on the market right now. It’s also $200 a month if you want the full experience. Goose, an open-source agent built by Jack Dorsey’s Block, does essentially the same thing—for free. And unlike Claude Code, it doesn’t care which AI model you use.
The project, which started life as an internal Block tool called goose, has quietly accumulated 42,531 GitHub stars and 4,294 forks. This week, it moved from Block’s GitHub organization to the Agentic AI Foundation (AAIF) at the Linux Foundation—a governance shift that puts it on the same footing as other major open-source infrastructure projects. The codebase remains Apache 2.0 licensed.
Where Claude Code lives exclusively in your terminal and requires Anthropic’s Max plan ($200/month) for unrestricted access, Goose ships as a native desktop app for macOS, Linux, and Windows, plus a full CLI and API. It’s built in Rust for speed and supports providers ranging from Anthropic and OpenAI to Google, Ollama, Azure, and Bedrock. KDnuggets wrote a detailed tutorial calling it a tool that “goes beyond code suggestions to autonomously execute tasks.”
Why Goose Is Gaining Traction Against Claude Code
The economics are hard to ignore. Claude Code on Anthropic’s Pro plan ($20/month) comes with usage limits that users are hitting faster than expected. The Max plan removes those limits but costs $200 monthly. Goose is free—you just bring your own API keys, which means you can plug in an existing ChatGPT, Claude, or Gemini subscription via ACP providers and pay nothing extra for the agent layer.
But cost isn’t the only story. Goose connects to over 70 extensions through the Model Context Protocol—the same standard Anthropic created to link AI agents to external tools. That means Goose can interact with your file system, run terminal commands, call APIs, and handle multi-step development workflows autonomously. It doesn’t just suggest code—it writes it, tests it, debugs it, and deploys it.
Block itself deployed AI agents across 12,000 employees in just eight weeks, according to The New Stack. Goose was part of that rollout. The company’s VP of Engineering, Angie Jones, described it as a tool that moves “beyond code suggestions to automate meaningful engineering tasks.”
The move to the Linux Foundation matters for adoption. Enterprise teams are wary of betting on tools controlled by a single company—especially one whose primary business is payments, not developer tools. The AAIF governance model means Goose now has independent stewardship, contribution guidelines, and a path toward community-driven development.
Goose is available as a desktop download at goose-docs.ai or via CLI install with a single curl command. The project’s governance details are on the GitHub repository.