Claude Code: Anthropic's developer agent, explained

Claude Code is Anthropic's tool that turns Claude into a true developer agent: it works directly in your terminal, IDEs, the browser and the app, right next to your code.

What is Claude Code?

It is a command-line agent (also integrated into IDEs) that reads your codebase, understands your project, edits files, runs commands, runs tests and automates whole engineering tasks — always under your approval.

What it can do

Implement a feature end to end, fix bugs, write tests, refactor, explain an unfamiliar repo, review code, automate repetitive tasks. It can also drive external tools via the MCP protocol.

How to get started

Claude Code is installed and launched from the terminal, authenticating with your Anthropic account (Pro/Max plans or an API key). You describe the goal in natural language and approve its actions as it goes.

Follow Claude Code updates

Claude Code evolves fast. Find its announcements, tips and discussions in the Code category of Claude News, and the product context in our complete guide to Claude.

Frequently asked questions

Is Claude Code free?

Claude Code is used with an Anthropic plan (Pro/Max) or a usage-billed API key. Exact terms evolve; check the product news for up-to-date details.

What's the difference between Claude and Claude Code?

Claude is the assistant (chat, writing, analysis). Claude Code is the specialized agent that acts directly on your code in the terminal and IDEs.

Can Claude Code run commands on its own?

It proposes and runs actions (commands, file edits, tests) under your control, with permission guardrails.

Does Claude Code work with my IDE?

Yes — besides the terminal, it integrates with IDEs (VS Code, JetBrains), the browser and the desktop app.

Claude News is an independent publication, not affiliated with Anthropic.