Glossary · Anthropic

Claude Code

Claude Code is Anthropic's terminal-based AI coding agent — reads your repo, runs commands, edits files, and ships PRs. Definition, pricing, and use cases.

By Kadin Nestler · May 28, 2026 · Updated May 28, 2026
On this page

What Claude Code does

You point Claude Code at a directory and give it a task in natural language. It explores the repo, plans changes, runs commands (tests, builds, lints), edits files, and reports back. Unlike copy-paste-from-chat workflows, the agent has direct file access and can iterate against test results without a human round-trip. It uses MCP for extensibility — you can plug in custom tools, knowledge bases, or APIs.

How it differs from competitors

  • Cursor — IDE-integrated, GUI-first, multi-model. Claude Code is terminal-first, single-model.
  • GitHub Copilot — autocomplete-style suggestions in the editor. Claude Code is agentic, multi-file.
  • Cline / Continue — open-source VS Code extensions. Claude Code is Anthropic-first and tightly model-tuned.
  • Devin (Cognition) — cloud-hosted agent. Claude Code runs locally with your code on your machine.

Pricing

Claude Code is available on Anthropic's Claude Max plan ($100-$200/mo per user) with included usage allotments, or via the API with pay-per-token billing on top of any Claude model. Power users running long agentic sessions typically prefer the Max plan because the token math gets expensive on hourly use. Anthropic explicitly markets Max as the right tier for Claude Code workloads.

When SMBs care about Claude Code

Most SMBs do not directly use Claude Code — it is a developer tool. But if you hire an AI agency, the productivity multiplier matters: agencies running on Claude Code can ship 30-day-build sprints in 7-14 days. Ask your agency what coding agent they use and how much of the build is human vs agent. The honest ones will tell you it is 60-80% agent in 2026 and that the price you pay reflects that leverage.

What it means for your business

Claude Code is the leverage point that makes a one-or-two-person AI agency competitive with a 20-person dev shop. If your vendor is not using a frontier coding agent, you are paying for typing they should not be doing.

  • Claude Agent SDK — The Claude Agent SDK is Anthropic's open-source framework for building production agents on top of Claude. Definition, capabilities, and when to use it.
  • Claude Opus — Claude Opus is Anthropic's most capable model, tuned for deep reasoning, long context, and agentic coding. Definition, pricing, and when to use it.
  • Claude Sonnet — Claude Sonnet is Anthropic's balanced model — strong reasoning, lower cost than Opus, faster latency. Definition, pricing, and use cases.
  • AI Agent — An AI agent is an LLM-driven program that uses tools to complete tasks autonomously. Definition, architecture, and real SMB examples.
  • Model Context Protocol (MCP) — MCP is an open standard for connecting AI models to tools, data, and APIs. Definition, ecosystem, and why it matters for AI interoperability.
← Back to the full glossary