Claude Code Guide 2026: What It Is, Install, Pricing, Skills & How It Compares

What Is Claude Code?

Claude Code is an AI-powered coding assistant from Anthropic, the company behind the Claude family of AI models. Unlike a simple autocomplete tool, it is agentic: you describe a goal in plain language and it explores your project, plans an approach, edits multiple files, runs commands and tests, and can even create git commits and pull requests.

According to Anthropic’s docs, typical uses include:

• Building features and fixing bugs from a plain-language description or an error message.
• Automating tedious work such as writing tests, fixing lint errors, resolving merge conflicts, updating dependencies and writing release notes.
• Working with git: staging changes, writing commit messages, creating branches and opening pull requests.
• Connecting your tools through the open Model Context Protocol (MCP), for example Jira, Slack or Google Drive.
• Scripting and CI: run it non-interactively (claude -p), pipe logs into it, or use it in GitHub Actions and GitLab CI/CD.

Where You Can Use Claude Code

Claude Code runs on several surfaces, all connected to the same underlying engine, so your repository’s CLAUDE.md files, settings and MCP servers work across all of them.

How to Install Claude Code

Anthropic recommends the native install, which updates itself in the background. Run the command for your system, then open a project and start Claude Code.

curl -fsSL https://claude.ai/install.sh | bash

irm https://claude.ai/install.ps1 | iex

curl -fsSL https://claude.ai/install.cmd -o install.cmd && install.cmd && del install.cmd

brew install --cask claude-code

winget install Anthropic.ClaudeCode

Start Your First Session

cd your-project
claude

You’ll be prompted to log in on first use. If the ANTHROPIC_API_KEY environment variable is set, Claude Code skips the login prompt and asks you to approve the key instead.

Claude Code in VS Code, Cursor & JetBrains

Claude Code for VS Code

1. Open the Extensions view (Ctrl+Shift+X / Cmd+Shift+X) and search for Claude Code (extension ID anthropic.claude-code).
2. Install it, then open the Command Palette (Ctrl+Shift+P / Cmd+Shift+P), type Claude Code, and choose Open in New Tab.

The extension adds inline diffs, @-mentions, plan review and conversation history inside your editor. Claude Code in Cursor: the same extension can be installed in Cursor.

Claude Code vs. the VS Code Extension

They are not competing products: the extension is a surface of Claude Code. The terminal CLI is the most full-featured; the extension gives you a visual, editor-native experience. Because they share the same engine, your CLAUDE.md, settings and MCP servers work in both.

JetBrains

Install the Claude Code plugin from the JetBrains Marketplace and restart your IDE. It requires the Claude Code CLI to be installed separately.

How to Use Claude Code

A typical session follows a simple loop: you describe a goal, Claude Code explores and proposes a plan, makes changes, checks its work, and you review.

Practical Examples

# Ask for a task
claude "write tests for the auth module, run them, and fix any failures"

# Let it handle git
claude "commit my changes with a descriptive message"

# Use it in a pipeline (non-interactive)
git diff main --name-only | claude -p "review these changed files for security issues"

Habits That Make It Work Better

• Be specific. “Add input validation to the login function in auth.ts” beats “improve this codebase.”
• Use plan mode (press Shift+Tab) for bigger changes, so Claude proposes an approach before touching files.
• Give it a way to verify: tests, expected output, or a screenshot.
• Add a CLAUDE.md to your project root with coding standards, architecture notes and preferred libraries; it is read at the start of every session. Claude Code can also read an existing AGENTS.md.
• Review every diff. Claude Code is powerful, but you stay responsible for what gets committed.

Claude Code Skills, Plugins, Subagents, Hooks & MCP

Claude Code is extensible. Several layers let you customize it, from a single instructions file to fully packaged plugins.

Claude Code Skills

Skills (also searched as “Claude skills” or “Claude Code skills”) are reusable instruction packages. Each is a folder with a SKILL.md file. Claude loads a skill automatically when its description matches what you’re doing, or you can run it directly with a slash command (a folder named deploy becomes /deploy).

A minimal skill, adapted from Anthropic’s docs, that summarizes uncommitted changes:

---
description: Summarizes uncommitted changes and flags anything risky. Use when the user asks what changed, wants a commit message, or asks to review their diff.
---

## Current changes

!`git diff HEAD`

## Instructions

Summarize the changes above in two or three bullet points, then list any risks you notice.

Claude Code Plugins

A plugin is a directory of skills, agents, hooks and MCP servers that Claude Code installs as one unit, usually from a marketplace. Run /plugin in a terminal session and browse the Discover tab; Anthropic’s official marketplace is added the first time you start an interactive session. Plugins can be installed for you (user scope), for everyone in a repository (project scope), or for you in one repo (local scope).

Subagents, Hooks & MCP in One Line Each

• Subagents: spawn multiple Claude Code agents that work on different parts of a task; a lead agent coordinates.
• Hooks: run shell commands before or after Claude Code actions, such as auto-formatting after every edit.
• MCP: the open Model Context Protocol lets Claude Code connect to external tools and data sources.

Claude Code Pricing & Cost

Questions like claude code pricing, claude code cost and claude subscription all come down to two billing styles: a subscription with included usage, or pay-per-token through the API or a cloud provider.

Subscription Plans (Anthropic Pricing Page)

On subscription plans, usage draws from a per-seat allowance that resets on a rolling five-hour window plus a weekly window, and it is shared with Claude chat and Cowork. If you hit the limit, usage credits let you keep working (manage them with /usage-credits).

Pay-Per-Token: API, Console & Cloud Providers

You can also authenticate with a Claude Console account (or set ANTHROPIC_API_KEY), or run Claude Code through Amazon Bedrock, Google Cloud or Microsoft Foundry. In these setups you pay per token consumed. Anthropic’s cost guide reports that across enterprise deployments the average is about $13 per developer per active day and $150–250 per developer per month, with costs below $30 per active day for 90% of users. Your own numbers will vary with model choice, codebase size and how you use it.

How to Keep Claude Code Costs Down

• Clear between tasks with /clear; stale context costs tokens on every message.
• Match the model to the job with /model: Sonnet handles most coding tasks for less than Opus; reserve Opus for hard reasoning.
• Use plan mode to avoid expensive rework, and write specific prompts.
• Delegate verbose work (tests, logs) to subagents so only a summary returns.
• Keep CLAUDE.md short (under about 200 lines) and move specialized workflows into skills.
• Check usage with /usage and /context, and disable unused MCP servers and plugins.

Claude Code vs Claude Cowork vs Claude Chat

“Claude Cowork vs Claude Code” is a common question. Public reviews describe Claude Cowork as a desktop agent for knowledge work (documents, research, spreadsheets, presentations) for people who don’t want a command line, while Claude Code is for people who write software. Both sit alongside regular Claude chat.

Claude Code vs Codex, OpenCode, Antigravity & Other Agents

The AI coding-agent space moves quickly. This table summarizes what each tool is designed to be, based on vendor pages and public comparisons; verify current features and pricing before you commit.

How to Choose

Because they use different models and philosophies, results vary by task. The most reliable way to choose is to run the same real task in two tools and compare the diffs, the cost and the time you spent reviewing.

Claude Code Docs, Updates & Troubleshooting

• Claude Code docs / documentation: Anthropic’s official docs at code.claude.com/docs include a quickstart, best practices, CLI reference, hooks, MCP, subagents, and third-party provider guides (Amazon Bedrock, Google Cloud, Microsoft Foundry).
• Claude Code news: see the official changelog for release notes. Run claude --version to check your version.
• Claude status: if requests fail unexpectedly, check Anthropic’s status page for an ongoing incident before debugging locally.
• Install problems: the docs include an install troubleshooting guide that maps common errors (such as a 403 or a curl error) to fixes.
• Usage-limit messages: “session limit” and “weekly limit” messages refer to your plan allowance; the message shows when it resets, and usage credits can extend it.

Sources

• Claude Code Docs: Overview
• Claude Code Docs: Skills
• Claude Code Docs: Plugins
• Claude Code Docs: Manage costs effectively
• Anthropic: Claude pricing
• DataCamp: Claude Cowork vs Claude Code
• SSOJet: AI Coding Agents Compared in 2026
• Kommunicate: Claude Code vs Codex vs Antigravity
• MindStudio: Hermes Agent vs Claude Code vs OpenClaw

This is an independent guide and is not affiliated with or endorsed by Anthropic. Claude and Claude Code are trademarks of Anthropic. Prices, plan features and competitor descriptions come from public sources as of September 2026 and can change; always confirm on the vendor’s official pages.