What Is Claude Code?
A terminal-native agent that reads files, edits them, runs tests, and explains itself — Anthropic's flagship CLI agent.
Claude Code is a terminal-based AI coding agent from Anthropic. It lives in your shell, reads files you point it at, edits them with your permission, runs tests, and explains every step. It is not a chatbot pasted into a sidebar; it is a process that operates on your machine alongside you.
The first thing to internalize about Claude Code is that the unit of work is not a token or a line — it is a feature. You describe the change you want; the agent reads, plans, edits, runs, observes, and adjusts. You review the diff. That four-beat loop — plan, act, observe, adjust — is the entire game.
Three model variants matter in 2026: Opus 4.7 (deepest reasoning, 1M context), Opus 4.6 (faster, slightly cheaper), and Sonnet 4.6 (fast, cost-efficient for routine work). The router can pick automatically; you can also pin a model for a task. Knowing which model is on the other side of your prompt is part of being a fluent operator.