Cursor vs Claude Code

If you are leaving Lovable, either one can take over your app. Our step-by-step guides for both have 6 to 19 steps. So the question is what you pay, which models you get for that money and how much the agent does without asking you.

That changed this month. On 12 October 2026 Cursor became part of The X Subscription, one plan for Grok, Cursor and X Premium. On the new plans Claude and GPT are not included. You can still use them in Cursor from Premium up, but you pay API list price from the first request, on top of the plan. Most comparisons in Google today were written before that.

We read the official docs and pricing pages of both on 26 October 2026. We did not run our own benchmark. Prices change often, so check the pricing pages before you pay.

Short answer

• Pick Cursor if you want an editor where you see every file, Cursor models are good enough for you or you are still on an old Cursor plan.
• Pick Claude Code if you want Claude models for a fixed monthly price, or you would rather give tasks to an agent than work in an editor.

Side by side

Prices checked on 26 October 2026, in US dollars, without tax. For Cursor these are the plans of The X Subscription that every new customer gets.

What you pay

Claude Code starts at Pro, $20 a month, or $17 a month if you pay $200 for the year. There is no free option. Max starts at $100 a month. Claude models are included, so the price stays the same until you hit a limit.

Cursor has more plans, and the details matter:

• Free: Agent and Chat with limited usage.
• Plus, $8 a month: Cursor models only (Grok 4.7, Grok 4.6, Grok 4.5 and Composer 2.5). No Claude, no GPT and no own API keys.
• Premium, $30 a month: 4x the Plus usage. Third-party models like Claude become available, at API list price from the first request, paid from prepaid credit. Your own API keys work too.
• Super ($100) and Ultra ($200): more usage for Cursor models. Claude costs the same as on Premium.

What does Claude cost in Cursor? Cursor's price table lists Claude Sonnet 5.5 at $2 per million input tokens and $10 per million output tokens, and Claude Opus 5.5 at $4 and $20. Those are the same numbers as on Anthropic's own pricing page. An agent sends your files and the conversation again on every step, so input tokens add up fast.

So if you want Claude, Claude Code is the cheaper way to get it: in Cursor you pay $30 for Premium before the first Claude request. Cursor is cheaper only if Cursor models do the job for you. Then Plus at $8 is the cheapest paid plan of the two.

GPT in Cursor may not stay either. OpenAI says on its help center that it plans to end its contract to supply models to Cursor, with a proposed transition period that ends on 12 November 2026. The date is not final, and Cursor may stop access sooner.

Already paying for Cursor?

Your plan keeps working. The old Pro, Pro+ and Ultra plans keep their price and features unless you change or cancel them. Moving to The X Subscription is optional, but it is permanent. Cursor's help page says it "can't be undone, and Support can't undo it for you." If you use Claude in Cursor, look at what you spend on it before you switch.

What happens when you hit a limit

This matters more than the monthly price. You notice it in the middle of a migration.

Cursor gives you one usage pool that refreshes every week. It is shared with the agent in the editor, the CLI, Cloud Agents, Grok Bot and the Grok apps. When it runs out, you wait for the reset, pay from prepaid credit or upgrade. You see what is left in the editor settings and under Usage at grok.com/dashboard. Cursor does not say how big the pool is, only how many times the Plus pool each plan gets. Claude and other third-party models do not use the pool at all. They come out of prepaid credit, so set the monthly limit for additional usage before your first Claude request.

Claude Code on Pro and Max has a five-hour limit and a weekly limit. Both are shared with Claude chat, so if you also use Claude for writing, that comes out of the same budget. When you hit a limit you can wait, move to a bigger plan or turn on usage credits, billed at API rates. The /status command in Claude Code shows what is left.

How much it does without asking

Cursor has three Run Modes, in Settings > Agents > Approvals & Execution:

• Auto-review: commands on your allowlist run straight away, other shell commands run in a sandbox (the project folder, network blocked except for common package managers) and the rest goes to a classifier.
• Allowlist: only what is on your list runs without asking. With an empty list it asks every time.
• Run Everything: no prompts, no sandbox and no classifier.

Cursor calls Auto-review the safest useful setup for most people, and in the same docs says it "is not a security boundary". The classifier can allow calls you would block.

Claude Code starts in auto mode from version 2.1.283, in the terminal and in VS Code. A second model reviews actions instead of you. Manual mode asks before most actions that edit files, run shell commands or reach the network. Press Shift+Tab in the terminal to switch, or start with claude --permission-mode manual. When you pick "Yes, and don't ask again" for a command, Claude Code saves a rule in .claude/settings.local.json and stops asking for it in that repository.

For a Lovable app this matters. Your Supabase project is live, with real users in it. For the first days, use Manual mode in Claude Code or Allowlist in Cursor, at least for anything that touches the database. And make a backup before the agent changes anything. Back up the database of a Lovable app shows how.

Editor or conversation

Coming from Lovable, this is the difference you feel on day one. Cursor is an editor. You have the file tree, every file opens in a tab and you read what the agent changed before you keep it. Claude Code in the terminal is a conversation. You give it a task and read what it did.

Claude Code is not only a terminal, though. The desktop app shows diffs visually. The extension for VS Code and Cursor shows each proposed edit side by side in Manual mode, before you accept it.

If you have never opened a code editor, Cursor will feel like a big step. We wrote more about that in Lovable vs Cursor.

Notes for the agent

Cursor reads AGENTS.md, a plain Markdown file in the project. For more control it has project rules in .cursor/rules. They must be .mdc files: a plain .md file there is ignored. Type /create-rule in Agent and it writes one for you. Cursor recommends keeping each rule under 500 lines.

Claude Code reads CLAUDE.md at the start of every session, and /init writes a first version from your code. If the repository has an AGENTS.md and no CLAUDE.md, Claude Code reads the AGENTS.md instead.

So one AGENTS.md works for both, and your notes do not lock you into either tool. If you wrote project knowledge in Lovable, Take your Lovable Knowledge to CLAUDE.md or AGENTS.md shows how to bring it along.

Which writes better code?

Most comparisons you will find tested Cursor with Claude inside it. On the new plans that costs extra. What you get for the plan price is Grok 4.7, Grok 4.6, Grok 4.5 and Composer 2.5. So for a new customer, Cursor vs Claude Code is now mostly Cursor models vs Claude, unless you also pay for Claude in Cursor.

We have not tested them against each other on a Lovable app, and we are not going to quote benchmarks for models that change every few weeks. For a typical Lovable app (React, Supabase, a few edge functions), give one real task to the model you would actually pay for and compare the result and the bill. Your code is a normal Git repository, so switching later costs you an install and a sign-in.

Can you use both?

Yes, and it can be the cheapest setup if you want Claude. Claude Code has an official extension for Cursor. Use Cursor as your editor and run Claude Code in it with your Claude Pro plan. You get the Cursor file view and Claude models without paying API prices for them.

If you do this, keep your notes in AGENTS.md, commit often and do not let both agents edit at the same time. If Lovable is still connected to the same repository, read Keep Lovable in sync when you edit in another tool first.

Getting started after Lovable

• Connect your Lovable project to GitHub. See Export a Lovable project to GitHub.
• Get the repository onto your computer and run it. See Run a Lovable project on your computer.
• Install the tool. Cursor: download the editor from cursor.com/download and open the project folder. Claude Code: run the install command below, then claude in the project folder. It asks you to log in on first use. Or open the Claude desktop app, sign in and click the Code tab.
• Give the agent its notes: an AGENTS.md for Cursor, or run /init in Claude Code to write a CLAUDE.md.
• Pick a safe mode before the first real task: Allowlist in Cursor, Manual in Claude Code.

Claude Code on macOS, Linux or WSL:

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

Prefer Cursor in the terminal? On macOS, Linux or WSL install the Cursor CLI, then run agent in the project folder:

curl https://cursor.com/install -fsS | bash

The full step-by-step guides go further: database, logins, edge functions, keys, hosting and the domain. They ask a few questions about your app and show only the steps you need. Still deciding whether to leave Lovable at all? Start with Lovable vs Cursor or Should you leave Lovable for Claude Code?

Pick Cursor if, pick Claude Code if

Pick Cursor if you want to see every file in an editor, Cursor models do the job for your app or you are on an old Cursor plan. Start on Plus at $8 if you only need Cursor models. Go for Premium only if you also plan to pay for Claude or GPT by usage.

Pick Claude Code if Claude is the model you want and you want a fixed monthly price for it, or you are happy to give tasks to an agent and check the result. Start on Pro at $20. Go for Max if Pro runs out too often. And if you still want an editor, run Claude Code inside Cursor.

Sources

1. Cursor: Pricing
2. Cursor docs: Models & Pricing
3. Cursor help: What is The X Subscription?
4. Cursor docs: API keys
5. Cursor docs: Run Modes
6. Cursor docs: Rules
7. Cursor docs: Cursor CLI
8. xAI: Introducing The X Subscription
9. xAI docs: The X Subscription pricing
10. xAI docs: Eligibility
11. OpenAI help: Using OpenAI models in Cursor
12. Claude: Pricing
13. Claude help: Use Claude Code with your Pro or Max plan
14. Claude Code docs: Overview
15. Claude Code docs: Use Claude Code in VS Code
16. Claude Code docs: Permission modes
17. Claude Code docs: Configure permissions
18. Claude Code docs: How Claude remembers your project

• Lovable to Claude Code
• Lovable to Cursor
• Lovable vs Cursor
• Should you leave Lovable for Claude Code?
• Take your Lovable Knowledge to CLAUDE.md or AGENTS.md
• Back up the database of a Lovable app
• Lovable pricing