GitHub Copilot vs Codex in 2026: which to pay for, and when to use both

Disclosure: Enjoy is our product, and it runs Codex, not GitHub Copilot’s own agents. We checked every fact below against GitHub’s, OpenAI’s and OpenCode’s own pages on November 3, 2026, and we say where you don’t need Enjoy at all.

GitHub Copilot vs Codex, compared

GitHub Copilot and Codex both run OpenAI’s GPT-6 models now, and they sell them in opposite ways. Copilot is an assistant in your IDE that suggests code as you type, with chat, an agent mode, a command line, a desktop app and agents in GitHub’s cloud, and models from several labs billed by the token from a monthly allowance of AI credits. Codex is OpenAI’s coding agent: you describe a change, and it reads your project, edits files and runs commands until it’s done, in the ChatGPT desktop app, a terminal, your editor or OpenAI’s cloud, on a ChatGPT plan with five-hour and weekly limits. Copilot can even run Codex itself, though GitHub lists no GPT-6 model for it.

Option	Best for	Price, checked November 3, 2026	Works in	Models
GitHub Copilot	Suggestions as you type, and an agent, in almost any IDE	Free; Pro $10 a month; Pro+ $39; Max $100; Business $19 a seat	VS Code, Visual Studio, JetBrains IDEs, Eclipse, Xcode; a CLI and the Copilot app	GPT, Claude, Gemini, Grok, Kimi and Microsoft’s; GPT-6.1 Sol from Pro+
Codex	Handing whole tasks to OpenAI’s GPT-6 models, on a ChatGPT plan	In every ChatGPT plan: Free and Go ($8 a month) with GPT-6 Luna in the desktop app; Plus $20; Pro $100, $200 or $500	The ChatGPT app on Mac and Windows, Linux in preview; a CLI; VS Code and its forks, JetBrains IDEs, Xcode	OpenAI’s GPT-6 models; GPT-6.1 Sol from Plus
Both, side by side	Copilot’s suggestions, with Codex’s agent work on your ChatGPT plan	ChatGPT Plus from $20 a month, plus Copilot from Free	VS Code	Copilot’s models in Copilot; GPT-6 in Codex
Codex inside Copilot	Handing GitHub issues to Codex on one Copilot bill	AI credits from your Copilot plan, plus GitHub Actions minutes; in public preview	GitHub.com, GitHub Mobile, VS Code	GPT-5.3-Codex and GPT-5.4 nano
Enjoy	Codex beside Claude Code, Grok Build and other labs’ agents	A monthly subscription on top of your plans, free for 7 days (pricing)	Mac, Windows, Linux	Claude Code, Codex, Grok Build, Antigravity, Muse Code, OpenCode

Prices are each company’s US list prices per month, before tax; GitHub’s plans page shows monthly prices only. Both sell to teams too: Copilot Business is $19 a seat a month and Enterprise $39, and Codex’s pricing covers ChatGPT’s Business and Enterprise plans.

How the two differ

	GitHub Copilot	Codex
What it is	An assistant in your IDE, with chat, an agent mode, a CLI, the Copilot app and a cloud agent	A coding agent, in the ChatGPT desktop app, a terminal, editor extensions and the cloud
Suggestions as you type	Unlimited on paid plans; 2,000 completions a month on Free	None; you give it a task
OpenAI’s models	GPT-6.1 Terra and GPT-6 Luna from Pro; GPT-6.1 Sol and GPT-6 Astra from Pro+; billed per token from AI credits	GPT-6.1 Sol, GPT-6.1 Terra and GPT-6 Astra in the plan from Plus ($20); GPT-6 Luna on Free and Go
Other labs’ models	Claude, Gemini, Grok, Kimi and Microsoft’s MAI-Code	None
Included usage	AI credits each month: 1,500 on Pro, worth $15	Five-hour and weekly limits, with no five-hour limit on Pro; shared with ChatGPT Work, and with ChatGPT from December
When it runs out	Upgrade, wait for the first of the month, or pay for additional usage	Wait for the reset, or buy credits
On GitHub	The Copilot cloud agent and code review, from Pro, billed in AI credits and Actions minutes	@codex review and @codex tasks on connected repositories, from Plus
Its command line	Copilot CLI, under a license to install and run it	The Codex CLI, open source under Apache 2.0
Project instructions	.github/copilot-instructions.md, AGENTS.md, or a root CLAUDE.md	AGENTS.md

GitHub Copilot

Best for: suggestions as you type in whichever IDE you use, with GPT, Claude and other labs’ models on one bill.

Copilot’s feature matrix lists code completion, chat and agent mode in VS Code, Visual Studio, JetBrains IDEs, Eclipse and Xcode, and code completion in Neovim. Beyond the editor, the Copilot CLI installs with npm, WinGet, Homebrew or a script, and the Copilot app is “a desktop application for agent-driven development” for macOS, Linux and Windows, “available for all Copilot plans.” Since GitHub Universe, “Copilot is now one connected experience in the cloud,” and “By default, every session runs in a secure cloud sandbox.” Its model list has OpenAI’s GPT-6.1 Sol and Terra, GPT-6 Astra, Sol and Luna, and Codex models, beside Claude, Gemini, Grok, Kimi and Microsoft’s MAI-Code.

Price: GitHub’s plans start with Free, with 2,000 completions a month, “access to models through auto model selection only” and, per its pricing page, “Limited chat and agent usage,” and Copilot Student is “Available to verified students” for free. Pro is $10 a month, with 1,500 AI credits a month, 1,000 base and 500 flex. Pro+ is $39, with 7,000 credits, and Max is $100, with 20,000. “1 AI credit = $0.01 USD,” so Pro’s allowance is $15 of chat and agent use, Pro+’s $70 and Max’s $200. “Code completions and next edit suggestions are not billed in AI credits and remain unlimited for all paid plans.”

Watch for: GPT-6 bills by the token, and its strongest models start on Pro+. “The cost of an interaction depends on two things: the model and the number of tokens consumed,” says its models and pricing page, which puts GPT-6.1 Sol at $2 per million input tokens and $10 per million output, GPT-6.1 Terra at $1 and $5, and GPT-6 Astra at $10 and $50, each more past 272,000 input tokens. Its model table marks GPT-6.1 Sol, GPT-6 Sol and GPT-6 Astra “Not included” on Pro. Per its billing guide, “Included AI credits do not carry over between months,” the allowance resets “at 00:00:00 UTC on the first day of each calendar month,” and “additional usage may be capped.” The pricing page adds that “Flex allotments may change over time.” And cloud sessions cost compute too: “Billing for cloud sandbox usage begins in November 2026,” per GitHub’s Universe changelog. Our memo on Universe has the details.

Codex

Best for: handing whole tasks to OpenAI’s newest models, on the ChatGPT plan you already pay for.

OpenAI’s desktop app is where Codex runs on your computer: “Run projects in parallel, work with files, use your computer, and keep long-running work moving from one desktop workspace.” It’s for macOS and Windows, and on Linux it “is available in preview.” On Windows, “It runs natively on Windows using PowerShell and the Windows sandbox,” per its Windows guide, or in WSL 2. The Codex CLI “is a coding agent from OpenAI that runs locally on your computer,” open source under Apache 2.0. Its IDE extension serves “VS Code and compatible editors,” while “Xcode and JetBrains IDEs provide their own integrations,” and Codex Cloud takes “coding tasks to work through in the cloud.” On GitHub, mention @codex review on a pull request and “Codex posts a review on the pull request, just like a teammate would,” flagging “only P0 and P1 issues”; any other @codex comment starts a task.

Price: Codex’s pricing gives Free, and Go at $8 a month, “GPT-6 Luna at Standard speed in the desktop app, subject to rollout.” Plus, at $20, adds “Codex on the web, in the CLI, in the IDE extension, and on iOS,” “Cloud-based integrations like automatic code review and Slack integration,” and “GPT-6.1 Sol and GPT-6 Luna.” Pro comes in three tiers, at $100, $200 and $500 a month, and “Pro plans currently have no five-hour limit.” OpenAI estimates that Plus allows 15 to 160 local messages every five hours with GPT-6.1 Sol, and 5 to 45 with GPT-6 Astra. Our ChatGPT Plus vs Pro guide covers which to buy.

Watch for: OpenAI says its counts “are not fixed message limits,” and “Weekly limits may also apply.” Codex shares them with ChatGPT Work, and from December with chat too: “Starting in December, Plus, Pro, and Business plans’ usage allowance will be shared across ChatGPT and Codex,” says its help center, with no size given yet. Our memo on it has what’s known. Free and Go get one model, in one place: GPT-6 Luna, in the desktop app. Codex doesn’t suggest code as you type, so if you still write much of the code yourself, you’ll want an editor’s suggestions beside it. And OpenAI’s IDE page names VS Code and compatible editors, Xcode and JetBrains IDEs, not Visual Studio or Eclipse.

Both: Codex beside Copilot in VS Code

Best for: keeping Copilot’s suggestions as you type, with Codex’s agent work on your ChatGPT plan instead of your AI credits.

You don’t have to choose. Codex’s IDE extension works “beside your code”: “Bring open files and selections into the prompt, review edits in place, and hand off longer work without breaking your flow.” OpenAI’s sign-in docs offer “Sign in with ChatGPT for subscription access” or “Sign in with an API key for usage-based access,” so signed in with ChatGPT, Codex’s work comes from your plan, and Copilot keeps its credits for chat and its own agent.

One instruction file can serve both. GitHub’s docs say “You can create one or more AGENTS.md files, stored anywhere within the repository,” and OpenAI’s docs say “Codex reads AGENTS.md files before doing any work.” Codex’s GitHub reviews read it too: add a “Code Review Rules” section, and “Codex applies the root and more-specific guidance that covers each changed file.”

Price: ChatGPT Plus, at $20 a month, or Pro, plus Copilot, from Free, or Pro at $10. Codex’s pricing lists the IDE extension from Plus; for Free and Go it lists only the desktop app. Copilot Pro and ChatGPT Plus together come to $30 a month, $9 less than Copilot Pro+, the plan where Copilot’s own GPT-6.1 Sol starts.

Watch for: two agents and two allowances in one editor. Copilot’s agent mode still bills AI credits when you use it, and Codex counts against your ChatGPT limits, so it’s on you to send each job to the one you mean. Codex’s extension also has a “Sign in with Copilot” option, whose usage is “subject to GitHub rate limits and billing” instead: that’s the next option.

Codex inside Copilot

Best for: handing GitHub issues and pull requests to Codex on one Copilot bill.

Copilot runs OpenAI’s agent itself. GitHub’s docs say “The OpenAI Codex coding agent and the VS Code OpenAI Codex integration use the Codex SDK and can be powered by your existing Copilot subscription.” You start it from a repository’s Agents tab, by assigning it an issue, by mentioning it in a pull request comment, from GitHub Mobile, or in VS Code, where you “Start a new session in the chat view, or delegate an existing session to a different agent.” In Codex’s own VS Code extension, the “Sign in with Copilot” option “is only available to GitHub Copilot Pro+ and Copilot Max subscribers.”

Price: “Coding agents consume GitHub Actions minutes and AI credits,” and “Each agent session consumes AI credits based on the model used and the number of tokens processed.” GitHub’s plans page lists third-party agents on Pro, Pro+ and Max, while its pricing page’s comparison table lists delegating to them from Pro+, so check your plan before you count on it.

Watch for: the models. GitHub lists Auto, GPT-5.3-Codex and GPT-5.4 nano, so none of the GPT-6 models a ChatGPT plan runs in Codex is there, and “Auto in OpenAI Codex coding agent does not leverage Copilot auto model selection.” The integration “is currently in public preview,” and it has to be turned on in your Copilot policies first: allowing it “will install openai code agent,” a GitHub App. Every session draws on the same AI credits as the rest of Copilot.

Enjoy

Best for: Codex with other labs’ agents in one conversation, when you’d rather describe a change than type it.

Enjoy is a desktop app for Mac, Windows and Linux that runs Claude Code, Codex, Grok Build, Antigravity, Muse Code, and local models through OpenCode, each signed in with the plan you already have. It doesn’t run Copilot’s agents; it’s another way to run Codex, and it adds:

• Another agent, mid-conversation. Hand a job from Codex to Claude Code or Grok Build: “The new agent gets the conversation so far, including replies from other agents, before it works on your message.”
• Limits handled. When Codex runs out, Enjoy switches to another of your accounts with usage left, or waits for the reset and carries on. When OpenAI gives your account a free reset, Enjoy shows it beside your usage.
• Copilot’s models, through OpenCode. OpenCode uses the providers you set up in it, and its homepage says you can “Log in with GitHub to use your Copilot account.”
• Your history, brought along. Open your Codex projects in Enjoy and import their conversations.

Price: Enjoy is one monthly subscription, and the first seven days are free, with a card to start. It includes your phone and any browser and your other computers, and each teammate you invite needs a subscription of their own. See pricing.

Watch for: Enjoy isn’t an editor in Copilot’s sense. Its code editor doesn’t suggest code as you type, and it doesn’t plug into VS Code or JetBrains IDEs: the agents write the code. It doesn’t add GPT to a Copilot plan or put Codex on your AI credits: each agent runs on its own company’s plan, and OpenCode on the providers you give it. Waiting out a limit needs “Your computer must be on, with Enjoy open, when the usage resets,” and scheduled workflows also run only while it’s awake with Enjoy open. Enjoy isn’t open source, and it’s a monthly subscription on top of your AI plans once the free week ends.

When to stay with the one you have

Stay with Copilot if you write much of the code yourself and suggestions as you type are what you’d miss, or if you work in Visual Studio or Eclipse, which OpenAI’s IDE page doesn’t name. Pro, at $10 a month, is half ChatGPT Plus’s price, includes GPT-6.1 Terra and GPT-6 Luna, and keeps completions unlimited without touching your credits. It’s also the simpler bill if you want Claude and Gemini beside GPT.

Stay with Codex if you already hand it whole tasks and rarely miss an editor’s suggestions. Plus, at $20, runs GPT-6.1 Sol, which Copilot keeps for Pro+ at $39 and bills by the token; its limits are counted over five hours and a week, not a calendar month; and the Codex CLI is open source, where Copilot CLI’s license grants the right “to install and run copies” of it.

Add Codex to Copilot, rather than leaving, when your AI credits run out before the month does on agent work: the extension moves that work onto a ChatGPT plan without leaving VS Code. Move from Codex to Copilot, or add it, when you want suggestions as you type, or Claude and Gemini on the same bill.

How we compared

We read GitHub’s pricing, plans, billing, models and pricing, supported models, feature matrix, third-party agents, Copilot CLI, Copilot app and custom instructions pages, its Universe changelog and Copilot CLI’s license; OpenAI’s Codex pricing, ChatGPT pricing and usage limits pages, and its desktop app, Linux, Windows, CLI, IDE extension, sign-in, cloud, GitHub review and AGENTS.md docs; and OpenCode’s homepage, all on November 3, 2026, and every one is linked below. We quote each company’s own words for its limits and leave out what it hasn’t documented, such as how many tasks a month of AI credits covers or how big ChatGPT’s shared allowance will be. Plans and models change often, so we recheck this guide against its sources at least once a month, and the date at the foot of the page says when we last did.