Claude Code

Anthropic's agentic coding tool that lives in your terminal

claude.com · Free tier; Pro $20/mo; Max $100+/mo; API $3/M input, $15/M output

What it is

Claude Code is Anthropic's AI coding agent. You describe what you want in plain English and it plans, edits files across your codebase, runs commands, and iterates until things work. Available as a CLI, desktop app, web app, and IDE extensions.

Why it's good for vibe coding

It handles multi-file, multi-step builds — scaffolding whole apps, wiring auth and payments, debugging its own output. It's the tool we recommend first for building complete clones from the starter kits on this site.

Capabilities

  • Agentic coding: plans and executes multi-file changes, runs shell commands, fixes its own errors, iterates until tests pass
  • Surfaces: terminal CLI, desktop app (Mac/Windows), web (claude.ai/code), VS Code and JetBrains extensions
  • Handles: full-app scaffolding, refactors, debugging, git workflows (commits, PRs via gh), test-driven loops
  • Context: reads whole repos on demand; CLAUDE.md project memory; MCP tool integrations; hooks; subagents for parallel work
  • Models: Opus/Sonnet class models; best-in-class for long multi-step builds
  • Pricing: included in Claude Pro ($20/mo) and Max plans, or API pay-as-you-go
  • Best for: building a complete clone end-to-end from prompts; the default recommendation
  • Not for: people who want a visual editor-first workflow (see Cursor)