Review

Claude Code Review: What It Does, What It Costs, Who It Fits

Published Updated 4 min readBy Vibe Coding Intel editorial

Compiled from official docs and public sources, checked .

Verdict: Claude Code is the pick if you want an agent that works on your whole repository from the terminal, can be scripted into CI, and also runs in VS Code, JetBrains, a desktop app and the browser. It costs $20 a month on Claude Pro at the entry level, and a daily user should expect Max ($100) or Ultra ($200), or API billing that Anthropic puts at about $13 per developer per active day across enterprise deployments. Pick a graphical editor such as Cursor instead if you want to choose between model vendors, or want the agent to keep billing overage instead of pausing at a limit.

Where it runs

Anthropic’s overview docs, read 1 November 2026, list five surfaces that share one engine, so a project’s CLAUDE.md, settings and MCP servers carry across all of them:

SurfaceWhat the docs say it adds
Terminal (CLI)The full-featured version: edit files, run commands, manage the project from the command line
VS Code (and Cursor)Inline diffs, @-mentions, plan review and conversation history in the editor
JetBrainsPlugin for IntelliJ IDEA, PyCharm, WebStorm and others, with diff viewing
Desktop appmacOS and Windows (Linux in beta); visual diffs, parallel sessions, scheduled tasks, cloud sessions
Webclaude.ai/code; long-running and parallel tasks on repositories you do not have locally

What it does

  • Repository-wide changes. Plans an approach, writes code across files and verifies it; traces bugs from an error message to a fix.
  • Git work. Stages changes, writes commit messages, creates branches and opens pull requests; GitHub Actions and GitLab CI/CD integrations automate review and triage.
  • Scripting. Follows the Unix model: pipe logs or diffs into claude -p and chain it in CI.
  • Customisation. CLAUDE.md project instructions, shareable skills, hooks that run shell commands before or after actions, and MCP servers for outside tools. Our CLAUDE.md guide and MCP servers guide cover the two most used.
  • Parallel and scheduled work. Subagents coordinated by a lead agent, background sessions, and scheduled tasks that run in the cloud or on your machine.

It runs Anthropic’s Claude models only. We have not run a benchmark of our own and do not repeat vendor benchmark claims; Claude Code vs Cursor explains why cross-tool benchmark numbers do not compare.

What it costs

From claude.com/pricing, the Ultra help page and the API price list, read 1 November 2026:

RoutePriceLimits
Pro$20 a month, $17 billed annuallyFive-hour and weekly limits, shared with Claude chat
Max$100 a month5x Pro usage every five hours and every week
Ultra (formerly Max 20x)$200 a month, monthly only10x Pro usage a week; no five-hour limit for verified accounts
APIPer token, for example Sonnet 5.5 at $2 in and $10 out per millionNo allowance; every token billed

The pricing page also lists Fable, Anthropic’s top model, separately: on Pro it is paid with usage credits, and on Max it is included up to half the weekly limit; the help centre says Max and Ultra carry a separate weekly limit for it.

Anthropic’s help centre confirms the subscription limits are shared: activity in Claude and in Claude Code counts against the same allowance. When you hit a limit you can wait, upgrade, or turn on usage credits billed at API rates. As a real-usage anchor, Anthropic’s cost docs report about $13 per developer per active day and $150 to $250 a month across enterprise deployments. The full breakdown, with a worked example, is in Claude Code pricing explained.

Who should use it, and who should not

If you…Then
Work in the terminal and want to script the agent or run it in CIClaude Code fits
Want the same agent in VS Code, JetBrains, desktop and the browserClaude Code fits
Want to pick models from several vendors in one toolUse Cursor, or an open-source agent such as Cline
Prefer overage billing to a pause at the limitCursor bills on; Claude Code pauses unless usage credits are on
Are deciding between Anthropic and OpenAIRead Codex vs Claude Code

Method

Researched from Anthropic’s Claude Code documentation, pricing page, help centre and API price list, read 1 November 2026. We have not measured Claude Code’s output quality, speed or usage on a fixed task; costs are vendor prices and vendor-reported figures.

Common questions

How much does Claude Code cost?

It comes with Claude Pro ($20 a month, $17 billed annually), Max ($100) and Ultra ($200), or you pay per token through the API. Anthropic reports about $13 per developer per active day on API billing across enterprise deployments.

Is Claude Code only a terminal tool?

No. The terminal CLI is the full-featured version, but Anthropic also ships a VS Code extension (which also installs in Cursor), a JetBrains plugin, a desktop app and a web version at claude.ai/code.

What happened to Max 20x?

Anthropic renamed it Ultra. Existing Max 20x subscribers moved to Ultra automatically with no price change, per the Ultra help page.

Does Claude Code share limits with Claude chat?

Yes. On Pro, Max and Ultra, activity in Claude and Claude Code counts against the same limits.

Sources

Update log

  1. Rewritten with sources from Anthropic's Claude Code docs, pricing page and help centre, read 1 Nov 2026. Corrected the plans: Max 20x is now called Ultra ($200 a month); Max is $100 at 5x Pro. Removed unsourced claims about reasoning strength and fewer correction rounds. Added where it runs, what it automates, a cost table and a who-should-skip section.