GitHub Copilot Plans and Usage Limits 2026: What Each Tier Actually Gets You
· 10 min read

GitHub Copilot Plans and Usage Limits 2026: What Each Tier Actually Gets You

A plain-language guide to GitHub Copilot's tiers, what monthly AI credits cover, what stays unlimited, and how to choose the right plan for your workflow.

GitHub Copilot Plans and Usage Limits 2026: What Each Tier Actually Gets You

Last verified: 2026-10-18

PlanPrice (USD/month)Published limitsSource
Free$02,000 code completions/month; limited chat and agent usage; AI credits allowance exists but the number is not published; auto model selection only per usage-based billing docs; CLI and agent mode supportedGitHub Copilot plans
Pro$101,500 total monthly AI credits (1,000 base + 500 flex); unlimited code completionsGitHub Copilot plans, usage-based billing docs
Pro+$397,000 total monthly AI credits (3,900 base + 3,100 flex); unlimited code completionsGitHub Copilot plans, usage-based billing docs
Max$10020,000 total monthly AI credits (10,000 base + 10,000 flex); unlimited code completionsGitHub Copilot plans, usage-based billing docs
StudentPrice not publishedUnlimited code completions; AI credit allowance; auto model selection onlyusage-based billing docs
Plans-page dollar equivalentsPro $15 / Pro+ $70 / Max $200 total (base + flex shown as dollars)—GitHub Copilot plans

What GitHub Copilot plans exist and what do they cost?

GitHub Copilot has five individual plans: Free at $0, Pro at $10, Pro+ at $39, Max at $100 (all per month in USD on the GitHub Copilot plans page), and Student (price not published). The same page shows dollar equivalents of $15, $70, and $200 for Pro, Pro+, and Max.

Paid tiers are monthly subscriptions. Moving up mostly changes your monthly AI credit pool and which features you get, such as code review, the cloud agent, or delegating work to third-party agents, per the GitHub Copilot plans page. Code completions are unlimited on Pro, Pro+, and Max. Pricing on this page is what GitHub published on the fetch date, so treat it as the current reference point rather than a permanent guarantee.

What are AI credits and which features use them?

An AI credit is GitHub’s unit for AI work: 1 AI credit equals $0.01 USD, per the usage-based billing documentation. Chat, CLI, the cloud agent, Spaces, Spark, and third-party coding agents use credits. Code completions and next edit suggestions do not, and they are unlimited on all paid plans.

Conversations and agent tasks use credits because each one may involve multiple model calls, and the cost depends on which model answers and how many tokens it handles as input, output, or cached context. Agent mode and the cloud agent especially can burn through credits quickly because one task becomes many calls. Base credits are used first and the flex allotment applies after that, per the usage-based billing documentation.

How many credits does each plan include?

Pro includes 1,000 base credits plus 500 flex credits for 1,500 total monthly AI credits; Pro+ includes 3,900 base plus 3,100 flex for 7,000 total; and Max includes 10,000 base plus 10,000 flex for 20,000 total, per both the plans page and the usage-based billing documentation. The Free plan has an allowance, but the number is not published.

There’s a meaningful distinction between the two halves of your allowance. GitHub’s docs say base credits never change — they are the stable part of your plan. The flex allotment is explicitly variable: the docs describe it as designed to adapt, and the plans page states that flex allotments may change over time. In practice, that means the headline totals of 1,500, 7,000, and 20,000 credits are accurate today, but the flex portion should not be treated as a long-term guarantee. If you’re budgeting for a heavy agent workflow, plan around the base credits and treat flex as a bonus.

What happens when you run out of credits?

Base credits are used first, then the flex allotment applies automatically at the same rates, per the usage-based billing documentation. Unused credits are forfeited and the allowance resets to its full amount at 00:00:00 UTC on the first day of each calendar month. Pro, Pro+, and Max can buy more credits.

Two caveats worth knowing before you rely on overages. First, extra usage draws down a dollar budget you set at $0.01 per credit, and the docs say it may be capped; default cap values were not verified, so check the docs before enabling add-on purchases. Second, paid plans get a 10% discount on model costs, per the usage-based billing documentation, when you use auto model selection in Copilot Chat, Copilot CLI, the GitHub Copilot app, or the Copilot cloud agent. That discount applies to the model-cost side of usage, which helps stretch a fixed allowance further during long agent sessions.

What features differ between tiers?

Free ($0) includes agent mode, the CLI, and 2,000 completions a month but excludes code review, the cloud agent, and custom instructions. Pro ($10) adds those plus PR assignment. Pro+ ($39) and Max ($100) add third-party agent delegation to Claude and OpenAI Codex (Preview), per the GitHub Copilot plans page.

Model access is another differentiator that beginners often miss. On the fetch date, the plans page listed the same model set for every plan, covering models from Anthropic, OpenAI, Microsoft, Google, xAI, and Kimi, while the usage-based billing documentation says Free uses auto model selection only. That’s a snapshot: model lists change, and the exact per-plan model access should be confirmed on the plans page before you publish a specific claim about which model you’ll get. One more line item applies broadly: pooled, org-wide usage is marked as not included on the Free, Pro, Pro+, and Max individual plans. If you need shared team usage, that’s a separate conversation.

GitHub Copilot vs alternatives: which should you pick?

Copilot is the obvious fit if your daily life already happens inside GitHub and VS Code or another supported editor, and you want code completion plus chat and agent workflows bundled into one monthly subscription. The unlimited completions on paid tiers matter if you write code constantly, while credits cover the conversational and agent-heavy work.

Alternatives worth considering include other AI coding assistants and IDE plugins, which may suit you better if you want a different pricing model, a different editor ecosystem, or a tool deeply specialized in one workflow such as review or refactoring. Some developers also prefer to mix and match, using one assistant for completions and a separate one for long-running agent tasks. The honest trade-off is ecosystem versus flexibility: Copilot rewards you for staying in GitHub, while independent tools can be easier to move between environments. Evaluate alternatives on fit with your editor, your team’s workflow, and your comfort with metered usage — not on headline numbers alone.

FAQ

These answers cover the questions beginners ask most about Copilot credits, completions, and extra usage, based on GitHub’s official plans page and billing documentation. In short, included credits reset monthly and do not roll over, completions are not billed in credits, and paid plans can buy more credits.

Do Copilot credits carry over month to month? No. Included credits do not roll over; they reset at 00:00:00 UTC on the first day of each calendar month, per the usage-based billing documentation. Note that this reset is fixed to the calendar month and is not tied to the date you were billed, so a mid-month upgrade does not give you a fresh cycle aligned to your payment.

Are code completions billed against my credits? No. Code completions and next edit suggestions are not billed in AI credits and are unlimited on all paid plans, per the usage-based billing documentation. The Free plan is the exception, where the plans page lists 2,000 code completions per month rather than an unlimited allowance.

What is one AI credit actually worth? One AI credit equals $0.01 USD, per the usage-based billing documentation. However, the cost of a single chat message or agent task is not a fixed credit price: it depends on the model you use and the number of input, output, and cached tokens, and agent tasks involve multiple model calls.

Can I buy more credits if I run out? Yes — Pro, Pro+, and Max can purchase additional AI credits once the included allowance is used, according to the GitHub Copilot plans page. Extra usage draws down a dollar budget you set at $0.01 per credit, so a $10 budget covers 1,000 credits, and the usage-based billing documentation notes it may be capped.

What does the Free plan include? The Free plan at $0 includes 2,000 code completions per month, CLI access, and agent mode, with limited chat and agent usage and an AI credits allowance whose number is not published, per the GitHub Copilot plans page. Per the usage-based billing documentation, model access on Free is limited to auto model selection only, and per the plans page code review, the cloud agent, and custom instructions.md are excluded.

The Bottom Line

For most individual beginners, Pro at $10 per month is the value pick, and I’d rate it 8/10 on value. It includes unlimited code completions, code review, and the cloud agent, per the GitHub Copilot plans page, plus 1,500 monthly AI credits, per the usage-based billing documentation.

Move up to Pro+ at $39 or Max at $100 only if you genuinely live in agent workflows: those tiers raise the credit pool to 7,000 and 20,000 respectively, per the usage-based billing documentation, and add third-party agent delegation for Claude and OpenAI Codex (Preview), per the GitHub Copilot plans page. Stay on Free at $0 if you’re learning, experimenting, or light on chat usage; the plans page lists 2,000 monthly completions and limited chat and agent usage at no cost, so you can try it before you spend anything.

How This Guide Was Built

This guide is based on official documentation and pricing pages — we did not run the tool hands-on. Every price, credit figure, and feature limit in this guide comes from two official GitHub pages fetched on 2026-10-18, and nothing here comes from our own testing.

Every price, credit figure, and feature limit in this article comes from two official pages fetched on 2026-10-18: the GitHub Copilot plans page and the usage-based billing documentation for individuals. All citations are inline links to those sources — there are no numbered footnotes and no separate references list, because every claim can be checked at the point where it appears.

Not verified:

  • Default spend-cap values for additional usage.
  • Free plan’s numeric AI-credit allowance (not published).
  • Typical credit cost of one chat or agent task (the per-model pricing page at GitHub Copilot models and pricing was not fetched in this run).

We also did not verify whether premium-request-based limits from earlier Copilot billing still apply anywhere, and we make no per-model access claims beyond what the plans page listed on the fetch date. Where GitHub itself uses flexible wording — such as flex allotments “may change over time” — we have repeated that wording rather than translating it into a firmer promise. If you’re making a purchasing decision for a team or an organization, re-check the plans page on the day you subscribe.

📊 See how GitHub Copilot compares →

  • ToolBrain — tool reviews, LLM comparisons, and AI workflow guides
  • CodeIntel Log — code quality, debugging, and software engineering benchmarks

Cross-links automatically generated from None.

Back to all posts