Cursor vs Claude Code: Which One Should You Pay For on a Mac?

Most Cursor vs Claude Code comparisons still say both cost $20 a month. Since 12 October 2026 that is wrong for anyone signing up today. Cursor now sells The X Subscription, the same plan xAI sells for Grok. Its cheapest tier, Plus at $8, runs Grok models only. Claude only appears from Premium at $30, and every Claude request is billed on top, at Anthropic's API list price. Claude Pro still costs $20 and includes Claude Code.

That price change is the real story of this comparison, and none of the top-ranking pages we checked has caught up yet.

This guide compares the two tools on what they charge today and what each one is built to do. It shows how to run Claude Code inside Cursor with the Claude plan you already pay for, and what each one does on a Mac, down to the Xcode integration. Every price here was checked against the vendors' own pages on 18 October 2026.

The Key Takeaways

• Cursor changed its plans on 12 October 2026: new customers get The X Subscription, with Plus at $8 (Grok models only), Premium at $30, Super at $100 and Ultra at $200 a month.
• Claude costs extra in Cursor now: on Premium and above, Claude Opus 5.5 is billed at $4 per million input tokens and $20 per million output tokens, paid from prepaid credit, on top of the plan.
• Claude Code is included in every paid Claude plan: Pro at $20 ($17 a month billed yearly), Max at $100 or $200, with limits that reset every five hours plus a weekly cap.
• You can run both: the Claude Code extension installs in Cursor and signs in with your Claude plan, with no API key required.
• Old Cursor Pro still exists for existing subscribers at $20, with its own allowance for Claude and other third-party models. If you have it and use Claude, keep it.

Cursor vs Claude Code at a Glance

Cursor is an AI code editor you work inside, and Claude Code is Anthropic's coding agent that you hand tasks to. Cursor is built on the VS Code codebase and bundles models from several labs. Claude Code runs only Claude models, but it runs almost anywhere: in the terminal, in VS Code style editors (Cursor included), in the Claude desktop app and on the web. Here is how the two compare as of 18 October 2026.

What Changed on 12 October: Claude Now Costs Extra in Cursor

On 12 October 2026, Cursor stopped selling its old plans to new customers and moved them to The X Subscription. That is the bundle xAI launched the same day, covering Grok, Grok Bot, Cursor and X Premium. It follows SpaceX completing its acquisition of Cursor on 14 August.

For someone choosing between Cursor and Claude Code, the change that matters is simple: Claude is no longer part of any Cursor plan's included usage.

The New Cursor Plans

The X Subscription has five individual tiers. Free gives limited Agent and Chat requests. Plus costs $8 a month, and Cursor's pricing page states plainly that "Only Grok models are available on this plan". Premium costs $30 and adds 4x the usage of Plus, Cloud Agents, Bugbot code reviews, code hosting and third-party models. It also unlocks Grok Bot, which xAI describes as "AI teammates that get work done". Super costs $100 for 15x the usage of Plus, and Ultra costs $200 for 50x. Cursor's documentation lists yearly billing at $80, $300, $1,000 and $2,000 respectively.

The included pool covers Cursor's own models only: Grok 4.7, Grok 4.6, Grok 4.5 and Composer 2.5. It refreshes weekly and is shared with the Grok apps, so time spent in Grok Chat comes out of the same allowance as your coding. On Plus, even those models run at fixed settings, with Fast mode off and medium effort. For the full tier breakdown, including Teams, see our Cursor pricing guide.

What Claude Costs Inside Cursor

On Premium and above, Claude is billed separately at Anthropic's API list price, from the first request. Cursor's models and pricing documentation says third-party models like Claude, GPT and Gemini "aren't part of the included pool". You pay for them from prepaid credit that you top up yourself. Plus and Free cannot use Claude at all.

For Opus, Sonnet and Fable, the per-token rates match Anthropic's own price list exactly. Claude Opus 5.5 costs $4 per million input tokens and $20 per million output tokens. Sonnet 5.5 costs $2 and $10, and Fable 5.1, the top model, costs $10 and $50. Cursor lists Haiku 5.5 at a flat $0.10 and $0.50, which is Anthropic's rate for prompts up to 100K tokens. Cursor sells no discounted Claude bundle, so a Claude-heavy month in Cursor costs the plan price plus whatever your tokens add up to.

There is a second route. On Premium, Super and Ultra, and on legacy Pro, Pro+ and Ultra, Cursor lets you paste in your own Anthropic API key. Plus does not allow it. Anthropic then bills you directly per token, and those requests do not touch your Cursor pool. That is still API billing, though, not your Claude subscription.

If You Are Still on Cursor Pro at $20

Legacy Cursor Pro, Pro+ and Ultra keep working for existing customers. xAI's launch post says "nothing changes" unless you choose to switch. Cursor's documentation says the legacy plans keep two monthly usage pools, one for Cursor models and one for other models. That second pool is the allowance for Claude and GPT that new plans no longer include. If you already pay $20 for Pro and use Claude in Cursor every day, do not switch to the new plans without doing the maths first.

What Claude Code Costs on a Claude Plan

Claude Code is included in every paid Claude plan, and the cheapest is Claude Pro at $20 a month, or $17 a month billed yearly as $200 up front. Max costs $100 for 5x the usage of Pro, or $200 for 20x. The free Claude plan does not include Claude Code. Anthropic also sells API access per token, for anyone who would rather pay per request.

The limits work differently from Cursor's. Anthropic's pricing page says usage resets on "a rolling five-hour session window", and paid plans add a weekly limit on top. Claude Code shares that pool with your chats on web, desktop and mobile. When you hit the limit you can wait for the reset, upgrade, or switch on usage credits that bill at standard API rates. The full plan-by-plan detail, including Team seats, is in our Claude Code pricing guide. If you are torn between the two individual tiers, our Claude Pro vs Max comparison works through it.

One line in Anthropic's FAQ matters most here: Claude Code "shares the same usage limits as the rest of your plan". A long chat session eats into your coding time.

A Worked Monthly Example

A developer using Claude as much as Anthropic's average Claude Code user would pay about $180 to $280 a month for Claude inside Cursor, against a flat $20 to $200 on a Claude plan. The number comes from Anthropic's Claude Code cost documentation. Across enterprise deployments, it says, the average is "around $13 per developer per active day and $150-250 per developer per month" at API rates. It also says 90% of users stay below $30 per active day.

Use that as a ballpark for the same Claude models inside Cursor, where they bill at the identical API rates. That means roughly $150 to $250 a month in Claude usage, plus $30 for Cursor Premium. On a Claude plan, the same person pays a flat price instead: $20 for Pro, or $100 or $200 for Max if Pro's five-hour limit gets in the way. Cursor's agent is a different harness from Claude Code, so its token use will not match Anthropic's figure exactly. The direction of the gap will hold, though: for Claude-heavy work, a flat Claude plan beats paying per token.

The picture flips if you are happy with Grok. Cursor Plus at $8 gives you a working coding agent for less than half the price of Claude Pro, as long as you never need a Claude model.

Cursor vs Claude Code: How Each One Works

Cursor keeps you in the editor writing code with AI beside you. Claude Code takes a task and works through it on its own. That difference matters more than any model choice, because it decides how much of the work you still do yourself.

Cursor: An Editor You Code In

Cursor looks and behaves like VS Code, because it is built on the VS Code codebase, and it can import your VS Code settings and extensions in one step. You type, and Tab completion predicts the next edit. When you want more, Cursor's Agent edits across files while you watch the diffs. On Premium and above, Cloud Agents run tasks remotely, and Bugbot reviews pull requests. The appeal is control: you see every change as it happens, and you can switch between Grok, Claude, GPT and Gemini from one model picker.

Claude Code: An Agent You Delegate To

Claude Code is built to be given a job. You describe the task, and it reads the codebase, plans, edits several files, runs the tests and fixes what fails. A CLAUDE.md file in your project tells it your conventions at the start of every session, and it will read an existing AGENTS.md instead. Skills package workflows you repeat, hooks run your own scripts before or after its actions, and MCP connects it to tools like Jira or Google Drive. Because it is a command-line tool, you can also pipe logs into it or run it in CI. Before you let it run on its own, read our guide to Claude Code permissions and Auto mode.

Can You Use Claude Code Inside Cursor?

Yes. Anthropic publishes an "Install for Cursor" link for the Claude Code extension, and the extension signs in with your Claude account. Anthropic's VS Code extension documentation says any paid Claude plan (Pro, Max, Team or Enterprise) works, and that "no API key is required". Claude Code then runs in a panel inside Cursor and draws on your Claude plan's limits, not on Cursor's pool or credit.

This is the setup most people asking whether their Claude subscription works in Cursor are really after, and it is the cheapest way to get both tools. Keep Cursor on Free or Plus as your editor. Then do your Claude work through the Claude Code extension on a $20 Pro plan. For the price of Claude Pro plus Cursor Plus, $28 a month, you get the editor, a Grok agent, and Claude Code on Opus 5.5.

Cursor's own model picker is a different thing. It will not take your Claude subscription. It accepts an Anthropic API key, which Anthropic bills per token. If you pick Claude Opus 5.5 in Cursor's Agent, you pay per token either way, through Cursor's prepaid credit or through your own key. Cursor's documentation also notes that Tab completion always uses Cursor's built-in models, whichever key you add.

Cursor vs Claude Code on a Mac

Both tools run natively on a Mac, and both now plug into Xcode, which matters if you build for Apple platforms. The install paths are different, and so is how each one reaches Xcode.

Installing Each One

Cursor ships as a Mac app with separate downloads for Apple silicon (ARM64), Intel (x64) and a Universal build. Its command-line agent installs with one line: curl https://cursor.com/install -fsS | bash. Anthropic's setup page lists macOS 13 or later and 4 GB of RAM for Claude Code. It installs with curl -fsSL https://claude.ai/install.sh | bash or with Homebrew, using brew install --cask claude-code.

If you would rather avoid the terminal entirely, the Claude desktop app for Mac includes Claude Code in its Code tab. It runs several sessions side by side and can schedule recurring tasks. It needs a paid plan.

Xcode and Swift

Xcode 26.3 added agentic coding with Anthropic's Claude Agent built in, alongside OpenAI's Codex, according to Apple's February announcement. If you live in Xcode, you can use Claude without leaving it. Xcode 26.3 also opens its tools to any agent through MCP, and Cursor documents exactly how to connect. You add xcrun mcpbridge as an MCP server, and Cursor's agent can then read and edit files, trigger builds, run tests and capture SwiftUI previews. Xcode needs to be running with your project open, and Cursor's guide has you toggle Xcode Tools on under Settings, Intelligence, Model Context Protocol. Cursor lists a paid Cursor plan as a prerequisite. Because the bridge is plain MCP, Claude Code can connect to the same server with one command: claude mcp add --transport stdio xcode -s user -- xcrun mcpbridge.

Where Codex Fits

OpenAI's Codex is the third tool in most of these searches, and it is paid for the same way as Claude Code: through a chat subscription. ChatGPT Plus at $20 a month includes "Expanded Codex usage", and even the free ChatGPT plan has limited Codex access. Codex is also built into Xcode 26.3 next to Claude Agent. If you already pay for ChatGPT, Codex is the agent you have already bought. If you pay for Claude, it is Claude Code. Cursor is the only one of the three that sells you an editor first and the models second.

Which One Should You Pay For?

Pay for a Claude plan if most of your AI coding runs on Claude models. Pay for Cursor if you want an AI editor and are happy with Grok or with paying per token.

Most developers who already use one end up adding the other in its cheapest form.

Choose Cursor When

Choose Cursor if you want to write the code yourself, with Tab completion and an agent working alongside you in the editor. It is also the right pick if you like switching between Grok, GPT, Gemini and Claude per task. If $8 a month for a Grok-powered agent is enough for you, nothing else in this comparison comes close on price. Premium earns its $30 if you need Cloud Agents and Bugbot reviews on your pull requests. And if you are on legacy Cursor Pro at $20 with its included allowance for other models, it is still one of the better deals in AI coding.

Choose Claude Code When

Choose Claude Code if you want to hand over whole tasks, like refactors, test suites or dependency upgrades, and review the result rather than every keystroke. It is also the better fit if you script your workflow or run agents in CI. And it suits anyone who would rather pay a flat $20 to $200 than watch a token meter. It does not force a new editor on you either. You can keep VS Code, JetBrains, Xcode or Cursor and run Claude Code inside it. If you are not paying yet, our guide to whether Claude Code is free covers the free routes that still exist.

For the Questions That Do Not Need an Agent

Not every coding question needs an agent with write access to your repo. Sometimes you only want to ask why an error happens, or check Claude's plan against a second opinion. That should not mean paying for a third subscription. Fello AI puts Claude, GPT, Grok, Gemini and DeepSeek in one native Mac app. Ask the same question to two models without picking a side or opening another tab.

The Verdict

If Claude is the model you want, Claude Code on a $20 Pro plan is now the cheaper way to get it. Before 12 October, Cursor vs Claude Code was a question of workflow at the same $20 price. It is now also a question of money, because Claude is the one thing a new Cursor plan does not include. The extension means you can still run Claude Code inside Cursor's editor. If you want Cursor's editor, keep it on Free or Plus and let your Claude plan do the Claude work. Pay Cursor $30 or more only for what Cursor alone does: Cloud Agents, Bugbot and one picker for models from Anthropic, OpenAI, Google and xAI.

Frequently Asked Questions