What is goose?
goose is an open-source AI agent that runs on a user's machine for software development and general-purpose tasks. goose supports research, writing, automation, and data analysis alongside coding. The project is part of the Agentic AI Foundation at the Linux Foundation, and its source is published under the Apache-2.0 license.
goose is built in Rust and offers a native desktop application, a command-line interface, and an API. These access options serve users working in desktop applications, terminal workflows, or software that embeds the agent. The repository describes goose as an extensible agent rather than only a code-suggestion tool.
What goose can do
goose combines task execution with model-provider choice and extension support. The repository names concrete coding actions as well as broader uses such as research and writing. Model Context Protocol connections extend the agent, while custom distributions let developers package providers, extensions, and branding together.
- goose can install, execute, edit, and test code.
- goose supports research, writing, automation, and data analysis.
- goose works with 15+ providers, including Anthropic, OpenAI, Google, and Ollama.
- goose connects to 70+ extensions through Model Context Protocol.
- goose provides an API for embedding the agent.
- goose supports custom distributions with preconfigured providers, extensions, and branding.
How goose works
goose runs on the user's machine through a native desktop app or CLI, with an API available for embedding. The desktop app supports macOS, Linux, and Windows. The repository provides desktop download links and a shell command that downloads and runs the CLI installer.
goose connects to model providers using API keys or existing Claude, ChatGPT, or Gemini subscriptions through ACP. Named providers include OpenRouter, Azure, and Bedrock in addition to Anthropic, OpenAI, Google, and Ollama. Extensions connect through Model Context Protocol, which the repository identifies as an open standard.
Who goose is for
goose serves developers who want an agent that can act on code, as well as users working on tasks outside software development. The repository explicitly names research, writing, automation, and data analysis. Desktop, terminal, and embedding options give these audiences different ways to access the same general-purpose agent.
- goose supports developers installing, executing, editing, and testing code.
- goose supports users carrying out research and writing tasks.
- goose supports users automating workflows or analyzing data.
- goose supports developers embedding an agent through an API.
goose pricing and access
goose is available as an open-source project under the Apache-2.0 license. Access options include downloading the desktop application, installing the CLI, and obtaining the source from the public repository. The supplied repository text does not list goose subscription plans or model-provider pricing, so those costs are not specified here.
goose strengths and setup requirements
goose offers cross-platform desktop access, terminal workflows, an embedding API, and a choice of model providers. Model Context Protocol support and custom distributions provide concrete extension options. Setup involves obtaining the application or CLI and connecting a provider through API keys or a supported existing subscription; running locally does not by itself describe where every connected model executes.
goose FAQ
What is goose?
goose is an open-source, general-purpose AI agent that runs on a user's machine. goose supports code installation, execution, editing, and testing, as well as research, writing, automation, and data analysis through desktop, CLI, and API access.
Is goose free?
goose is available as an open-source project under the Apache-2.0 license. The repository provides desktop downloads, CLI installation, and source access, but does not specify model-provider charges or goose subscription plans.
Which operating systems does goose support?
goose provides a native desktop app for macOS, Linux, and Windows. goose also offers a CLI for terminal workflows and an API for embedding the agent in other software.
Which AI providers does goose work with?
goose works with 15+ providers, including Anthropic, OpenAI, Google, Ollama, OpenRouter, Azure, and Bedrock. goose supports API keys and existing Claude, ChatGPT, or Gemini subscriptions through ACP.
Does goose have an API or integrations?
goose offers an API for embedding the agent and connects to 70+ extensions through the Model Context Protocol open standard. goose also supports custom distributions with preconfigured providers, extensions, and branding.
Who is goose for?
goose is for developers and users handling research, writing, automation, or data analysis. goose supports coding tasks beyond suggestions, including installation, execution, editing, and testing, with desktop and terminal access options.
From the community
No reviews yet
Be the first to review this tool. What you write helps the next person choose.




