Home › Rules › GitHub Copilot vs Cursor vs Claude Code

GitHub Copilot vs Cursor vs Claude Code: which one should a beginner use?

Prices and plans from GitHub's, Cursor's and Anthropic's own pricing pages and docs, checked Oct 30, 2026. Sources below.

• Comparison
• GitHub Copilot
• Cursor
• Claude Code
• Beginners
• Pricing

Short answer: GitHub Copilot has a free plan, and Copilot Pro is $10 a month with models from several AI companies. It works inside editors you may already have, like VS Code, and on the GitHub website. Cursor is its own code editor app with an AI agent built in. It has a free plan too, and paid plans start at $8 a month. Claude Code is Anthropic's coding agent. It has no free plan; it comes with Claude Pro ($20 a month). Already paying for Claude? Try Claude Code first. Already keep your code on GitHub, or want a low price with a choice of models? Try Copilot. Want one app where the editor and the AI live together? Try Cursor. All three read an AGENTS.md rules file, so you can switch later without rewriting your rules.

Comparing OpenAI's Codex too? See Cursor vs Claude Code vs Codex. This page swaps Codex for Copilot.

Side-by-side breakdown

Prices are per month in US dollars, before tax, from each company's own pricing page. All plans have usage limits; see limits and costs.

Pick this if...

Pick GitHub Copilot if

• You want a low price. Copilot Free is $0, and Copilot Pro is $10 a month with models from several companies (Anthropic, OpenAI, Google and others). Cursor's cheaper $8 Plus plan is Grok models only.
• You already use VS Code or another editor and don't want to switch. Copilot is an add-on inside the editor you have.
• Your code is on GitHub. On paid plans, you can hand Copilot a task on the GitHub website and it works in the cloud, with or without opening a pull request for you to review. GitHub calls this "Copilot in the cloud".
• You're a student. Verified students get Copilot Student for free. Verified teachers and maintainers of popular open source projects may get Copilot Pro for free.

Pick Cursor if

• You want one app where the files, the editor and the AI chat are all in one window, built around the agent.
• You want to try it for $0. The Free plan has "limited agent requests", which is enough to test it on a small project.
• You're fine with how Cursor's plans split models. Plus ($8) is Grok models only. On Premium ($30) and up, other companies' models cost extra at API list price, on top of the plan. Cursor's pricing FAQ recommends Premium for new AI users.

Pick Claude Code if

• You already pay for Claude Pro or Max. Claude Code is included, so trying it costs nothing extra.
• You're OK typing in a terminal, or you'd rather use the Code tab in the Claude desktop app. Anthropic's docs say the desktop app needs a paid plan.
• You like your editor. Claude Code has an extension for VS Code and Cursor, so you can run it inside either one.

Not sure yet? Copilot and Cursor both have $0 plans, so you can try both on the same small project before you pay anything. If you already have a Claude plan, add Claude Code to that test. Coming from a browser builder like Lovable? Our guide to moving a Lovable project to Cursor or Claude Code walks through the switch.

Limits and costs, in plain words

GitHub Copilot: AI credits

• What a credit is: GitHub measures chat and agent use in "GitHub AI Credits". One credit is worth one cent. A quick question to a small model costs a fraction of a credit. A long agent session on a top model across many files costs a lot more.
• What each plan includes: Pro ($10) includes 1,500 credits a month, Pro+ ($39) includes 7,000, and Max ($100) includes 20,000. Part of that is a "flex allotment" on top of the base amount (500 of Pro's 1,500), and GitHub says the flex part can change over time.
• What doesn't use credits: code completions (the gray text that appears as you type) and next edit suggestions. They're unlimited on paid plans. Copilot Free caps completions at 2,000 a month.
• What does: Copilot Chat, agent mode, the Copilot CLI, Copilot in the cloud and code review.
• When it resets: credits reset on the first day of each calendar month (at midnight UTC), no matter when you signed up. Unused credits don't carry over.
• If you run out: wait for the reset, switch to a cheaper model, upgrade (you pay only the price difference), or set a dollar budget for extra use. Extra credits are 1 cent each, so a $10 budget buys 1,000 more. Copilot Free can't buy extra credits.
• Free vs paid: Copilot Free picks the model for you (GitHub calls it auto model selection). It includes agent mode in the editor and the Copilot CLI, but not Copilot in the cloud. Pro lets you choose from a selection of models; Pro+ and Max add GitHub's premium models.

Cursor: a weekly pool

Every Cursor plan has a pool of AI usage shared between Cursor and Grok, and it refreshes weekly. Bigger plans get a bigger pool: Premium has 4x the usage of Plus, Super 15x, Ultra 50x. If you run out, you can buy top-offs or wait for the refresh. Models from other companies aren't part of the pool; they're billed separately at API list price on Premium and up.

Claude Code: your Claude plan's limits

Claude Code uses the usage that comes with your Claude plan. Anthropic's pricing page says "usage limits apply" and that Max gives you 5x or 20x more usage than Pro. Anthropic's setup docs say you can also use an Anthropic Console (API) account instead of a Claude plan.

Only Copilot puts a number on its monthly allowance (credits). Cursor and Claude describe theirs as pools and multiples. Whichever you pick, check the usage page in its settings to see what you have left.

What "agent" means in each one

All three can do more than suggest the next line. You describe a goal, and the AI works through the steps, edits files and runs commands while you watch and approve. Here's where that happens in each tool:

• Copilot: agent mode runs in your editor (VS Code, Visual Studio, JetBrains, Eclipse and Xcode) on every plan, including Free. On paid plans, Copilot in the cloud takes tasks from the GitHub website or GitHub Mobile and works on its own, and you can pick the same session up later in VS Code, the Copilot CLI or the Copilot app. GitHub's pricing page also lists handing tasks to other companies' agents, like Claude and OpenAI Codex, on Pro+ and Max (in preview).
• Cursor: the agent is the main way you work in the app. The agent command does the same in a terminal. Cloud agents, which keep running while you're away, start at Premium.
• Claude Code: it's an agent from the start. In the terminal you type what you want and approve the changes. The desktop app and the web version work the same way, with a visual view of the changes.

How each tool reads your rules file

A rules file is a plain text file in your project that tells the AI how you want things done, so you don't repeat yourself every chat. Here's what each tool reads, from its own docs:

One rules file for all three

Put your rules in AGENTS.md in the project root, and don't add a CLAUDE.md unless you need Claude-only notes. Cursor reads it, Claude Code reads it when there's no CLAUDE.md, and Copilot's agent mode in VS Code, Copilot in the cloud and the Copilot CLI read it too. Here's a short starter you can copy and fill in:

# AGENTS.md

## About this project
- A small web app for [who it's for] that lets them [the main thing it does].
- Built with [your stack, for example Next.js + Supabase].

## How to work
- Explain what you plan to change before you change it.
- Make one small change at a time, then tell me how to test it.
- Don't add new packages without asking me first.
- Never put API keys or passwords in the code. Use environment variables.
- Don't delete files or data without asking me first.

## Commands
- Install: npm install
- Run locally: npm run dev
- Tests: npm test

## When you finish a task
- Tell me what changed, how to check it, and suggest a commit message.

Copilot's code review in VS Code and Copilot Chat on the GitHub website only read .github/copilot-instructions.md. If you use those, add a short one. GitHub's docs don't show a way to point this file at AGENTS.md, so put the review rules you care about right in it:

# Copilot instructions

The main project rules are in AGENTS.md in the project root.

## When reviewing code
- Flag any API key, token or password that ends up in the code.
- Flag new packages that weren't there before.
- Keep comments short and in plain English.

More on which file wins: AGENTS.md vs CLAUDE.md and Cursor rules vs AGENTS.md. More starters: AGENTS.md examples. Or build a rules file for your exact project in the free rules generator.

How to try them for $0

1. Copilot Free: sign up for Copilot Free with your GitHub account, install the GitHub Copilot extension in VS Code and sign in to GitHub there. Or use Copilot on the GitHub website, with nothing to install. GitHub says Copilot Free is only for people who don't already get Copilot through an organization or enterprise.
2. Cursor Free: download Cursor for Mac, Windows or Linux, sign in, open a folder and start with a small task.
3. Claude Code: no $0 option on a Claude plan. If you already pay for Claude Pro or Max, install it and log in with that account.

Use the same small project for each, with the same AGENTS.md, so the test is fair. Ask each one to explain the project, make one small change, and tell you how to check it. Then go with the one whose answers you understood best.

Related

• Cursor vs Claude Code vs Codex for beginners (real prices)
• copilot-instructions.md examples
• AGENTS.md examples
• AGENTS.md vs CLAUDE.md (which file each tool reads)
• Cursor rules examples
• CLAUDE.md examples
• Cursor rules vs AGENTS.md
• Plan mode guide (6 tools + 6 plan prompts)
• Best AI tool to fix and refactor existing code

Sources

• GitHub: Copilot plans and pricing
• GitHub docs: Plans for GitHub Copilot
• GitHub docs: Usage-based billing for individuals (AI credits)
• GitHub docs: About GitHub Copilot
• GitHub docs: Quickstart for Copilot in your IDE
• GitHub docs: Where to use GitHub Copilot
• GitHub docs: Copilot on GitHub.com (Copilot in the cloud)
• GitHub docs: Adding repository custom instructions
• GitHub docs: Support for different types of custom instructions
• GitHub docs: Copilot for students
• Cursor: Pricing
• Cursor docs: Quickstart
• Cursor docs: CLI
• Cursor docs: Rules (AGENTS.md)
• Claude: Pricing
• Claude Code docs: Overview
• Claude Code docs: Setup (which accounts include Claude Code)
• Claude Code docs: Memory (CLAUDE.md and AGENTS.md)

All prices and plan details checked Oct 30, 2026 on GitHub's, Cursor's and Anthropic's own pricing and docs pages. Prices are in US dollars, before tax, and change often. If something here no longer matches, tell us at [email protected].