goose
Block-built open-source AI agent, now hosted by the Agentic AI Foundation
Overview
goose is a free, Apache 2.0 AI agent that runs locally as a desktop app, CLI or API, working with 15+ LLM providers or an existing AI subscription and extending through MCP. Block created it and contributed it to the Linux Foundation's Agentic AI Foundation, announced in December 2025; the repository and docs moved there in April 2026.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
Use cases
What you can produce with goose
- A local agent that edits files, runs shell commands, installs packages and executes tests
- One agent driven by any of 15+ LLM providers, a local Ollama model or an existing Claude, ChatGPT or Gemini subscription
- Connections to databases, browsers, GitHub, Google Drive and more through 70+ documented MCP extensions
- Repeatable workflows saved as shareable YAML recipes that can run in CI
- Parallel subagents for review, research or file processing
- An ACP server that lets Zed, JetBrains or VS Code talk to goose
goose is one of the most capable free agents you can run on your own machine: desktop, CLI and API from one Rust codebase, more than 15 model providers including local Ollama models, 70+ MCP extensions, shareable YAML recipes and built-in guards such as prompt-injection detection and a sandbox mode. Vendor-neutral governance under the Linux Foundation's Agentic AI Foundation lowers the risk of a single company changing course.
The trade-offs are the usual ones: quality tracks the model you plug in, you pay that provider, and there is no commercial support. Local execution suits teams with data-residency concerns. Strong value for developers comfortable with configuration.
What we found
goose is an open-source AI agent that runs on your own machine, with a desktop app for macOS, Linux and Windows, a command-line interface and an API. Block, the company behind Square and Cash App, introduced it as codename goose in January 2025. It is not limited to code: the project pitches it for research, writing, automation and data analysis as well as editing files, running shell commands and executing tests.
Stewardship has changed hands. On 9 December 2025 the Linux Foundation announced the Agentic AI Foundation (AAIF), with Block's goose, Anthropic's Model Context Protocol and OpenAI's AGENTS.md as its founding projects. The practical move followed on 7 April 2026, when the repository shifted from block/goose to the aaif-goose organisation and the documentation to goose-docs.ai, whose footer now carries the foundation's copyright. Earlier versions of this page dated the donation to April 2026; that was the migration, not the announcement. The licence is still Apache 2.0.
You bring the model. The project lists more than 15 providers, including Anthropic, OpenAI, Google, Ollama, OpenRouter, Azure and Bedrock, and it can drive an existing Claude, ChatGPT or Gemini subscription through the Agent Client Protocol (ACP). Extensions connect over MCP, with more than 70 documented. Recipes save a workflow as a shareable YAML file, subagents run tasks in parallel, and the safety features include prompt-injection detection, tool permission controls, a sandbox mode and an adversary reviewer that watches for unsafe actions.
Development is quick: v1.54.0 shipped on 8 October 2026, six days after v1.53.0. A 2.0 beta announced in April rebuilds the clients around ACP, with a new TypeScript terminal interface and a Tauri desktop app replacing the Electron one. The costs are the usual ones for a bring-your-own-model agent: output quality depends on the model you choose, you pay that provider directly, and help comes from GitHub issues and Discord rather than a support contract.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including goose's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with goose unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to goose directly →
Spotted something out of date? Suggest an update →
Related reading
Alternatives to goose
More in Code & Dev Tools
