Claude Code
AI coding agent that runs on macOS, Linux, and Windows, uses cloud model APIs, and edits code, runs tests, and opens pull requests.
Claude Code is an AI coding agent for developers who want to delegate work across a repository, from bug fixes to longer refactors and migrations. It runs locally on macOS, Linux, and Windows, but sends requests to cloud model APIs. It is proprietary software and requires authentication to a supported model service.
The agent searches the codebase to find relevant files and understand dependencies, so you don't have to assemble the context yourself. It can explain unfamiliar code, plan changes, edit files, and run commands and tests. For larger tasks, it follows imports across the repository and continues working when tests expose problems, while surfacing decisions that need your input.
Claude Code works in the terminal alongside existing command line tools, with editor extensions for VS Code and JetBrains. GitHub and GitLab connections let it read issues and open pull requests. MCP servers and plugins extend its capabilities with additional tools, custom commands, and agents. You can also access coding tasks through the browser, Claude desktop and mobile apps, or Slack, and work on several tasks at once.
The local terminal client talks directly to model APIs without a separate backend server or remote code index. Local execution doesn't mean offline inference: the models run in the cloud. Anthropic also collects usage feedback and associated conversation data.
