What to do when GitHub Copilot runs out of AI credits: 9 ways to keep coding in 2026

Disclosure: Enjoy is our product, and it runs Claude Code and Codex, not GitHub Copilot’s own agents. We checked every price and limit below against GitHub’s, Microsoft’s, Anthropic’s, OpenAI’s and OpenCode’s own pages on November 3, 2026, and we say where another way is the better pick, including the free ones that don’t need Enjoy.

The ways to keep coding, compared

Since June 1, every Copilot plan has billed chat and agents in GitHub AI Credits, at a cent a credit, by the model and the tokens used. All your models draw on one allowance, so switching to a cheaper one won’t help once it’s empty. GitHub’s billing page gives an individual plan three moves: upgrade, pay for more, or wait. On a company seat, the message one developer reports opens “You’ve reached your credit limit,” and the next move is your admin’s.

Way	Cost, checked November 3, 2026	How soon you’re back	Best for
Wait for the reset	Free	The first of next month, at 00:00 UTC	The end of a month, or when suggestions as you type are enough
Set a budget for additional usage	A cent a credit: $10 covers 1,000	Right away	Finishing the task in front of you
Upgrade your plan	The difference this month: Pro to Pro+ $29; Pro+ to Max $61	Right away	Running out most months
Ask your admin	Your company’s budget	When your admin approves it	A Business or Enterprise seat
Your own key or a local model	The provider’s API rates, or free on your own computer	Right away, in a new session	Staying in VS Code’s chat or Copilot CLI
Claude Code, beside Copilot	Claude Pro $20 a month, or $17 billed yearly; Max $100; Ultra $200	Right away, in a new conversation	A Claude plan you already pay for
Codex, beside Copilot	ChatGPT Plus $20 a month; Pro $100, $200 or $500	Right away, in a new conversation	A ChatGPT plan you already pay for
OpenCode, with free or local models	Free	Right away, in a new conversation	No bill and no plan to run out
Enjoy	A monthly subscription, free for 7 days, with your phone and other computers (pricing)	Right away, in a new conversation	Several labs’ agents, so the next limit doesn’t stop the work

Prices are each company’s US list prices, before tax. Copilot’s own Claude and Codex agents use your AI credits too, so they aren’t on this list; the Claude Code and Codex rows are those labs’ own extensions, on those labs’ plans.

First, check what ran out

A limit error isn’t always your credits. VS Code’s FAQ says a rate limit can appear while you still have credits, and that the fix is to wait a few minutes and retry. To see where you stand:

• On an individual plan, go to github.com/settings/billing and choose AI usage, which shows included credits used, additional usage, and what each model cost.
• On a company seat, open Copilot settings from your profile picture and look under Usage this cycle.
• In your IDE, select the Copilot icon in the status bar in VS Code, JetBrains IDEs and Eclipse, or the menu bar in Xcode. In Visual Studio, it’s Copilot Consumptions, under the icon at the top right.
• On an older annual Pro or Pro+ plan, you may still be on premium requests until the plan expires, so the credit figures below don’t apply to you.

Here’s what each plan includes, from GitHub’s plans page:

Plan	Price	AI credits a month	Worth
Copilot Free	Free	An allowance GitHub doesn’t size	Not stated
Copilot Student	Free, for verified students	An allowance GitHub doesn’t size	Not stated
Copilot Pro	$10 a month	1,500 (1,000 base, 500 flex)	$15
Copilot Pro+	$39 a month	7,000 (3,900 base, 3,100 flex)	$70
Copilot Max	$100 a month	20,000 (10,000 base, 10,000 flex)	$200
Copilot Business	$19 a seat a month	1,900 a user, pooled	$19
Copilot Enterprise	$39 a seat a month	3,900 a user, pooled	$39
Copilot access	$0 with GitHub Enterprise Cloud; Standalone $10 a user	None: billed from the first use	None

Wait for the reset

Best for: the last days of a month, or a stretch where suggestions as you type are all you need.

Your allowance comes back at 00:00 UTC on the first day of each calendar month, and that date doesn’t move with your billing date. In the US, that’s still the last day of the month. Until then, paid plans keep code completions and next edit suggestions, which aren’t billed in credits, and VS Code’s FAQ confirms that inline suggestions carry on when only your credits have run out.

Price: free.

Watch for: everything else billed in credits stops: chat, agent mode, Copilot CLI, the cloud agent, Spaces, Spark and third-party agents. Run out on the third and that’s four weeks. Credits don’t carry over, so a quiet month won’t pay for a busy one. Copilot Free has its own cap on suggestions, 2,000 a month, while Student’s are unlimited, and on Free your choices are a paid plan or next month.

Set a budget for additional usage

Best for: finishing the task in front of you without changing plans.

Open your billing overview, choose Budgets and alerts, then New budget, and pick Bundled AI credits budget to cover everything billed in credits. You can then keep working at the published per-token rates, with credits drawing down your budget in dollars.

Price: a cent a credit, so $10 covers 1,000. What a credit buys depends on the model. GitHub’s price list has GPT-6 Luna at $0.10 per million input tokens and $0.50 per million output, Claude Sonnet 5.5 and GPT-6.1 Sol at $2 and $10, and Claude Opus 5.5 at $4 and $20.

Watch for: tick Stop usage when budget limit is reached before you start. Without it, GitHub only emails you when you pass the budget, and usage goes on. Additional usage can also be capped: GitHub says it may limit individuals by usage patterns, billing history and verification status, and its billing page says you may have to pay off what you’ve used before you can continue. If you bought your plan in the iOS or Android app, you can’t buy credits through the app store, and extra usage goes to any payment method on your GitHub account; a bundled budget of $0 with the stop on is GitHub’s way to prevent that. GitHub’s pricing page doesn’t offer extra credits on Free.

Upgrade your plan

Best for: running out most months, not once in a while.

GitHub says you pay only the difference between plans, your usage so far counts inside the bigger allowance, and the extra credits are there at once. So Pro to Pro+ costs $29 and adds 5,500 credits, where $29 of additional usage buys 2,900. Pro+ to Max costs $61 and adds 13,000, against 6,100. Pro+ also opens the models GitHub’s plan table marks as not included on Pro, such as Claude Opus 5.5, GPT-6.1 Sol and Gemini 4 Argon.

Price: Pro $10 a month, Pro+ $39 and Max $100.

Watch for: the base credits match the price, and the rest is a flex allotment, which GitHub’s pricing page says may change. That’s 3,100 of Pro+’s 7,000 credits and 10,000 of Max’s 20,000. And the models a bigger plan opens spend faster: Claude Opus 5.5 costs twice what Claude Sonnet 5.5 does per token, so moving up to it halves what those credits buy.

Ask your admin

Best for: a Copilot Business or Enterprise seat, where the budget isn’t yours to change.

When you run out, GitHub shows a banner on GitHub.com offering to request more budget, and owners and billing managers can approve, adjust or deny it. Seats share a pool, which resets on the same date as individual plans. What happens past it is your company’s call: Microsoft’s Visual Studio docs say usage continues at published rates if additional usage is allowed, and otherwise it’s blocked until the next cycle or until an admin changes the policy or budget. A user-level budget can stop you even when the company’s budget has room.

Price: nothing to you, if it’s approved.

Watch for: suggestions stay available even at a $0 budget, so you won’t lose those. A personal plan isn’t a way around it: being assigned a seat cancels a personal Pro, Pro+ or Max plan, with a prorated refund. And check before you bring another agent to work: VS Code’s docs warn that organization policies can restrict these routes even when you have your own credentials.

Your own key or a local model

Best for: staying in VS Code’s chat or Copilot CLI, on someone else’s bill or none.

VS Code’s bring your own key connects a provider’s models to chat, and in Local sessions it works with no GitHub account or Copilot plan at all. Anthropic, OpenAI, Gemini and Azure are built in, and local Ollama models come through the official Ollama extension. Your provider’s billing and limits apply, and a Local session can hand off to VS Code’s Claude or Codex session with its history. Copilot CLI does much the same: it finds models Ollama is serving on your machine, takes other providers in a providers.json file, and on Windows PCs with the right hardware installs MAI Code 1.1 Flash, which GitHub says avoids spending credits. VS Code’s experimental on-device models show 0 credits per response, though they need a GitHub sign-in.

Price: the provider’s API rates, such as Anthropic’s $4 per million input tokens and $20 per million output for Claude Opus 5.5, and OpenAI’s $2 and $10 for GPT-6.1 Sol. A model on your own computer adds no bill.

Watch for: it covers chat and utility tasks only. Semantic search, inline suggestions and features that rely on embeddings still need a GitHub account. A chat subscription doesn’t fill this slot: OpenAI’s sign-in docs treat subscription access and API keys as separate routes. VS Code warns that local models may lack the safety filters of Copilot’s cloud models and may fall more easily for malicious instructions, so go carefully with automatic approvals. On a company seat, on-device models need your admin to turn on preview features.

Claude Code, beside Copilot

Best for: a Claude plan you already pay for, with Copilot’s suggestions still running.

Anthropic’s extension for VS Code and plugin for JetBrains IDEs run Claude Code next to Copilot. Any paid Claude plan works, with no API key, and it draws on Anthropic’s limits, not your credits. Your project’s instructions can come too: Copilot reads AGENTS.md files anywhere in a repository, and Claude Code can read them in place of CLAUDE.md.

Price: Claude’s pricing has Pro at $20 a month, or $17 a month billed yearly, Max at $100 and Ultra at $200. The free plan leaves Claude Code out.

Watch for: which Claude you start. VS Code’s built-in Claude session bills through Copilot unless you configure a Claude key, and the model you pick decides the bill turn by turn; for your plan, use Anthropic’s extension. Claude has limits too: Pro and Max have a five-hour session limit and a weekly limit, shared with the Claude apps, and our Claude Code usage limit guide covers what to do then. Anthropic’s help center names VS Code, its forks and JetBrains IDEs, not Visual Studio or Eclipse. You start a new conversation, and some features are only in its CLI.

Codex, beside Copilot

Best for: a ChatGPT plan you already pay for, and OpenAI’s newest models on it.

OpenAI’s extension serves VS Code and compatible editors, while Xcode and JetBrains IDEs have their own integrations. Sign in with ChatGPT and it uses your subscription, not your credits. It reads AGENTS.md too. To try it for nothing, VS Code’s experimental Codex integration accepts a free ChatGPT account in its Agents window, though it’s off under enterprise-managed settings.

Price: ChatGPT Plus is $20 a month, and Pro $100, $200 or $500. Codex’s pricing lists the IDE extension from Plus; Free and Go get GPT-6 Luna in the desktop app, subject to rollout.

Watch for: the extension’s Sign in with Copilot option, on Pro+ and Max, is subject to GitHub’s billing, so that’s your credits again. OpenAI estimates 15 to 160 local messages every five hours with GPT-6.1 Sol on Plus, says those aren’t fixed limits, and adds that weekly limits may apply. From December, that allowance is shared with ChatGPT, as our memo explains, and our Codex usage limit guide covers running out. OpenAI’s import lists Claude Code, Claude Cowork and Cursor, not Copilot, so you start a new conversation.

OpenCode, with free or local models

Best for: routine work with no bill and no plan to run out.

OpenCode is an open-source coding agent for your terminal, IDE or desktop. It says it’s free to use and comes with a set of free models, and it runs local models too. Our local model guide covers the setup and the memory it takes.

Price: free, with its free models or a model on your own computer. Connecting another provider can add costs.

Watch for: its free models are free for a limited time, and for several, the team is using that time to collect feedback and improve the model. Its GitHub login uses your Copilot account, which is the allowance you just ran out of. It starts a new conversation, and what runs well locally depends on how much memory your computer has.

Enjoy

Best for: several labs’ agents in one place, so when one runs out, another carries on.

Enjoy runs Claude Code, Codex, Grok Build, Antigravity, Muse Code, and local models through OpenCode, on Mac, Windows and Linux, on the plans you already have. It doesn’t run Copilot’s agents, so it won’t refill your credits; it’s for moving agent work to plans with limits of their own. What it adds:

• Another agent, mid-conversation. Hand a job from Claude Code to Codex: “The new agent gets the conversation so far, including replies from other agents, before it works on your message.”
• Limits handled. When an agent runs out, Enjoy switches to another of your accounts with usage left, or waits for the reset and carries on.
• Usage in view. It shows what each signed-in account has left and when it resets, for the agents that report it.

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. The agents are billed separately: Claude Code needs a paid Claude plan and Codex in Enjoy needs ChatGPT Plus or above, while Antigravity works with a free Google account and OpenCode with free or local models.

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 can’t import your Copilot chats, so you start a new conversation. OpenCode in Enjoy can log in to your Copilot account, but that’s the same credits. 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. If you only want Claude or GPT beside Copilot, the two extensions above don’t need Enjoy at all.

Why your credits ran out so fast

The size of the job matters more than the number of requests. GitHub names the main factors: long conversations, agent features that make several model calls per task, and the model you choose. In June, one student reported spending a plan’s 200 credits in about 10 to 20 agent requests. To make the next month last:

• Pick the model first, and keep it. GitHub’s prices run from $0.10 in and $0.50 out for GPT-6 Luna to $10 and $50 for GPT-6 Astra and Claude Fable 5.1, before any long-context rate. Switching mid-session makes the model rebuild its cache, and cached tokens typically cost a tenth of fresh input. Changing reasoning effort does the same.
• Take the discounts. On paid plans, auto model selection takes 10% off model costs in Copilot Chat and the cloud agent, and Copilot Fusion takes 30% off through November 30, 2026.
• Keep sessions short. VS Code’s guide says to start a new session for each new task, and Copilot CLI’s /compact shrinks a long one.
• Cap the big jobs. Copilot CLI’s session limit, in public preview, stops a session at a set number of credits, 30 at the least.
• Get warned early. Under Budgets and alerts, GitHub can email you at 90% and 100% of your included usage. Visual Studio warns at 75% by default.

When to stay on your plan and wait

If the first of the month is a few days off, waiting is free, and suggestions as you type keep going. If you run out once in a while, pay for that day with a small budget, with the stop ticked. Upgrade when it happens most months: once you’d spend more than the $29 difference, Pro+ is the better deal, for as long as its flex allotment holds.

And stay in Copilot, whichever agent you add, if suggestions are what you’d miss. Pro’s are unlimited, and Claude Code and Codex run beside them. Copilot Pro with Claude Pro or ChatGPT Plus comes to $30 a month, $9 less than Pro+, which is worth trying if your credits go on agent work. Leave only when credits are what stop you and suggestions aren’t what you use: our Copilot alternatives guide compares where to go.

How we compared

We read GitHub’s billing, plans, pricing, models, budgets, usage, session limit, third-party agent, custom instructions and Copilot CLI pages and its June changelog; VS Code’s FAQ, language model, harness and optimization docs; Microsoft’s Visual Studio docs; Anthropic’s pricing, help center and Claude Code docs; OpenAI’s pricing, help, sign-in, import and Codex docs; and OpenCode’s homepage and docs, all on November 3, 2026, and every one is linked below. The error message and the student’s report are those developers’ own accounts. We put each company’s limits in our own words, link the page that states them, and leave out what it hasn’t documented, such as the size of the Free and Student allowances or how many tasks a month of credits covers. Plans and limits 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.