The $200 Question

Anthropic's Claude Code has become the gold standard for AI-assisted software development, but its pricing has created an opening that Block, the financial technology company formerly known as Square, is aggressively exploiting. Claude Code's most capable tier costs 200 dollars per month and still imposes rate limits that reset every five hours. Goose, Block's open-source AI agent, offers nearly identical functionality for exactly zero dollars.

The comparison is stark. Anthropic's free plan provides no Claude Code access whatsoever. The Pro plan at 20 dollars per month limits users to between 10 and 40 prompts every five hours. The Max plans at 100 and 200 dollars per month offer 50 to 200 and 200 to 800 prompts respectively, plus access to Claude's most capable model. Goose, by contrast, runs entirely on a user's local machine with no subscription fees, no cloud dependency, and no rate limits.

What Goose Actually Does

Goose is a general-purpose AI agent that operates from the command line, much like Claude Code. It can read and modify files, execute shell commands, run tests, manage git operations, and orchestrate complex multi-step development workflows. It supports multiple AI model backends, meaning users can connect it to whichever language model they prefer, whether that is a cloud API like OpenAI or Anthropic, or a locally running open-source model.

The tool has gained significant traction since its release. As of January 2026, Goose boasts more than 26,100 stars on GitHub with 362 contributors and 102 releases. The latest version, 1.20.1, shipped on January 19, 2026, and the pace of development shows no signs of slowing.

The BYOM Advantage

Goose's bring-your-own-model architecture is both its greatest strength and its most important differentiator. Claude Code is locked to Anthropic's Claude models. Goose works with any model that exposes a compatible API. This means developers can choose their model based on task requirements: a smaller, faster model for routine code generation and a more capable model for complex architectural decisions.

For developers working on sensitive codebases, the ability to run everything locally with a local model is a significant security advantage. No code leaves the machine, no prompts are logged by a third-party service, and no proprietary algorithms are exposed to cloud providers.