Contact
Copilot

GitHub Copilot Licensing Guide

print
share

Summary

GitHub Copilot moved to usage-based billing on 1 June 2026, replacing premium requests with GitHub AI Credits. This guide covers all five paid plans, per-token model pricing, credit pooling and budgets, seat billing, and what each tier includes for procurement teams.

Updated: 23 June 2026


Key Takeaways

  • GitHub Copilot is not Microsoft 365 Copilot. Despite the shared "Copilot" branding, these are separate products with different contracts, pricing, and billing models.

  • There are five paid plans across two audiences. For individuals, GitHub Copilot offers Copilot Pro at $10 per month, Copilot Pro+ at $39 per month, and Copilot Max at $100 per month. For organisations, it offers Copilot Business at $19 per user per month and Copilot Enterprise at $39 per user per month. Copilot Enterprise requires GitHub Enterprise Cloud.

  • Usage-based billing is now live. Since 1 June 2026, GitHub Copilot measures usage in tokens and converts the cost into a virtual currency called GitHub AI Credits, where 1 credit equals $0.01 USD. Every plan includes a monthly credit allowance, and usage beyond that allowance is billed at per-token rates that vary by model.

  • Costs are variable and non-deterministic. Under token-based billing, the cost of an interaction depends on the model chosen and how many tokens it consumes. The same prompt can produce different token counts, so per-interaction costs cannot be predicted exactly.

  • Seat billing is consolidated. If a user holds Copilot seats in several organisations within the same enterprise, you pay for that user only once. Billing for a removed seat continues until the end of the current billing cycle.

  • Code completions remain unlimited. Code completions and next edit suggestions stay unlimited on all paid plans and are not metered under the new billing system.

  • Self-serve Business sign-ups are paused. Self-serve Copilot Business sign-ups for organisations on GitHub Free or GitHub Team have been temporarily paused since 22 April 2026. The separate individual sign-up pause that GitHub introduced on 20 April 2026 has since been lifted. Copilot Pro, Pro+, and Student are open to new sign-ups again, and Copilot Max is available as an upgrade for existing Copilot users.


1. What is GitHub Copilot?

GitHub Copilot is an AI coding assistant built into development environments. It helps developers write code faster by suggesting completions, generating functions, explaining code, and answering technical questions directly in their editor.

If Microsoft 365 Copilot is the productivity assistant for knowledge workers, GitHub Copilot is the equivalent for software developers.

What GitHub Copilot does:

  • Code completion: Copilot suggests lines or entire functions as developers type.

  • Chat: Developers can ask questions about their code, get explanations of errors, and generate tests.

  • Pull request summaries: Copilot automatically summarises code changes for reviewers.

  • Documentation generation: Copilot creates docstrings and README content from existing code.

  • Code review: Copilot suggests improvements and catches potential issues before merge.

  • Agentic coding: The Copilot cloud agent and agent mode run multi-step coding sessions, working across files, running tests, and iterating on their own output.

Where it works:

GitHub Copilot integrates with Visual Studio Code (VS Code), Visual Studio, JetBrains integrated development environments (IDEs) such as IntelliJ and PyCharm, Eclipse, Xcode, Neovim, and GitHub.com itself. All paid plans include Copilot Chat on GitHub.com as well as in IDEs.


2. How Does GitHub Copilot Differ from Microsoft 365 Copilot?

The shared "Copilot" name creates confusion, especially during procurement. GitHub Copilot and Microsoft 365 Copilot are distinct products, and treating them as one line item is a common source of budgeting errors, not least because Microsoft 365 Copilot pricing works on an entirely different basis. If you want a wider tour of the Copilot family before you buy, our guide to making sense of Microsoft Copilot sets out how the pieces fit together.

Aspect

GitHub Copilot

Microsoft 365 Copilot

Audience

Software developers

Knowledge workers

Environment

Code editors, GitHub.com

Word, Excel, Outlook, Teams

Data source

Code repositories, documentation

Microsoft Graph (emails, files, calendar)

Billing unit

GitHub AI Credits (token-based, since 1 June 2026)

Per-user licence (plus Copilot Credits for agents)

Base price

$10 to $100 (individuals), $19 or $39 per user (organisations)

$30 per user per month

Overage model

Per-token rates converted to AI Credits

Copilot Credits for agents

Contract vehicle

GitHub agreement, or a Microsoft agreement

Microsoft agreement, such as an Enterprise Agreement (EA) or Cloud Solution Provider (CSP)

Sources: GitHub Copilot Plans and Microsoft 365 Copilot Licensing.

Because internal requests often arrive as just "Copilot", clarifying which product a requester means is the first job of any procurement conversation.


🖐 Understand how Copilot fits into your wider Microsoft estate. Learn more: Microsoft Licensing Services for Enterprises.


3. GitHub Copilot Plans and Pricing

GitHub Copilot is sold to two distinct audiences. Individual developers buy Copilot Free, Copilot Pro, Copilot Pro+, or Copilot Max for their own use. Organisations buy Copilot Business or Copilot Enterprise and assign seats to their developers. The base prices below did not change when usage-based billing arrived. What changed is that the subscription now buys a monthly allowance of GitHub AI Credits rather than a fixed number of premium requests.

3.1 Plan Prices and Included Credits

Plan

Audience

Price per month

Included AI Credits

Copilot Free

Individual

No charge

A limited allowance, plus 2,000 code completions

Copilot Student

Verified students

No charge

An allowance, plus unlimited code completions

Copilot Pro

Individual

$10

1,500 (1,000 base plus 500 flex)

Copilot Pro+

Individual

$39

7,000 (3,900 base plus 3,100 flex)

Copilot Max

Individual

$100

20,000 (10,000 base plus 10,000 flex)

Copilot Business

Organisation

$19 per seat

1,900 per user

Copilot Enterprise

Organisation

$39 per seat

3,900 per user

For individual paid plans, GitHub splits the monthly allowance into two parts. The base credits match the subscription price and never change, and the flex allotment is an additional variable amount that GitHub adjusts as model pricing and efficiency evolve. Base credits are consumed first, then the flex allotment applies automatically at the same rates. The practical result is that a Copilot Pro subscriber has 1,500 credits to spend each month, not the 1,000 implied by the headline price.

For organisations, the included allowance is stated as a single figure per user, with 1,900 credits on Business and 3,900 on Enterprise. Existing Business and Enterprise customers receive a higher promotional allowance for the first three months of usage-based billing, which runs from 1 June to 1 September 2026. During that window, Business customers receive 3,000 credits per user and Enterprise customers receive 7,000. After the promotional period, the allowance returns to the standard 1,900 and 3,900.

Source: GitHub Copilot Plans, Usage-Based Billing for Organizations and Enterprises

3.2 Sign-up Availability

GitHub paused several sign-up routes in April 2026 to protect service reliability while it prepared for the billing change. The individual pause it introduced on 20 April 2026, covering Copilot Pro, Pro+, and Student, has since been lifted: GitHub's plans page again offers direct sign-up for those plans. One self-serve route remains closed. Since 22 April 2026, new self-serve Copilot Business sign-ups for organisations on GitHub Free or GitHub Team are temporarily paused, so an organisation on those plans cannot self-serve a new Business subscription and would go through GitHub sales instead. Copilot Max is available as an upgrade for users who already hold a Copilot plan.


4. Premium Requests, Tokens, and AI Credits

4.1 What Changed on 1 June 2026

Under the previous system, every chat or coding interaction cost a fixed number of "premium requests" regardless of how much work it involved. Each plan included a monthly allowance of premium requests, and a model's multiplier determined how many of those requests a single interaction consumed. GitHub's stated reason for the change was that a quick chat question and a multi-hour autonomous coding session could cost the same, making the premium request model unsustainable as agentic coding became the norm.

Since 1 June 2026, premium request units have been replaced by GitHub AI Credits. Usage is now measured in tokens and converted into credits according to the published per-token rate for each model.

4.2 How Tokens Become Credits

A token is the unit AI models use to process text. Token length varies, so a short common word such as "the" is one token, while a longer or less common word may be split into several. Every interaction consumes three types of token. Input tokens are the prompt and context you send to the model, output tokens are the response the model generates, and cached tokens are previously processed context that the model reuses, billed at a lower rate than fresh input.

GitHub prices each token according to the model used, then converts the total into GitHub AI Credits, where 1 credit equals $0.01 USD. Your plan includes a monthly allowance of credits, and once you exhaust it, most paid plans offer the option to purchase additional usage at the published rates. One exception applies. If you subscribe, or have subscribed, to a Copilot plan through GitHub Mobile on iOS or Android, the option to purchase additional AI credits is not available.

Because the same prompt can produce different token counts depending on context and model behaviour, usage-based billing is non-deterministic. You cannot know in advance exactly how many tokens, and therefore how many credits, a given request will consume.

4.3 No More Fallback to Cheaper Models

Under the old model, a user who exhausted their premium requests could fall back to a lower-cost model and keep working. That fallback no longer exists. Under usage-based billing, consumption is governed by available credits and by the budget controls administrators set. When the credits and any permitted additional usage run out, work stops until the next cycle or until an administrator raises the budget.


5. How GitHub AI Credits Work

5.1 What Consumes Credits

GitHub AI Credits apply to Copilot features that call AI models. These include Copilot Chat, the Copilot CLI (command-line interface), the Copilot cloud agent, Copilot Spaces, GitHub Spark, code review, and third-party coding agents. An agentic session is an autonomous, multi-step interaction where the AI writes code, runs tests, searches documentation, and iterates without waiting for human input at each step, so it can consume far more tokens than a single chat message.

Code completions and next edit suggestions are the exception. They are not billed in AI Credits, as section 8 explains in full.

5.2 Pooling, Budgets, and Hard Stops

For organisations, included credits are pooled at the billing entity level. An enterprise with 100 Copilot Business users holds a shared pool of 190,000 credits rather than 100 separate buckets, which lets a team of light users offset a heavy one. Adding licences mid-cycle increases the pool immediately, while removing licences does not shrink it until the start of the next cycle.

When the pooled credits are exhausted, what happens next depends on policy. If additional usage is allowed, work continues at the published per-credit rates and the extra spend is charged to the organisation. If additional usage is not allowed, Copilot is blocked until the next cycle refreshes the allowance.

Administrators control overage spending through budgets, which are set in US dollars and drawn down at the fixed rate of $0.01 per credit, so a $10 budget covers 1,000 credits. Budgets can be set at four levels: user, cost centre, enterprise spending limit, and organisation. A user-level budget caps an individual's draw from the shared pool, and a $0 user budget blocks that user immediately. There is no automatic fallback to a cheaper model when a budget is exhausted, so budgets are a genuine stop, not a slowdown.

One regional caveat applies. Data-resident and FedRAMP (Federal Risk and Authorization Management Program) compliant Copilot requests carry a 10% model multiplier increase. Organisations with data-residency requirements should therefore expect slightly higher consumption for the same work.


6. Legacy Model Multipliers for Annual Subscribers

Most users moved to token-based billing on 1 June 2026, but one group remains on the old system. Copilot Pro and Copilot Pro+ subscribers on existing annual plans remain on request-based billing until their plan expires, and for them the model multipliers increased on 1 June 2026. Under this legacy system, each model has a multiplier that determines how quickly it draws down the premium request allowance, so a model with a multiplier of 27 drains the allowance twenty-seven times faster than a model with a multiplier of 1.

These multipliers do not apply to usage-based billing. They apply only if you hold an annual Pro or Pro+ plan that has not yet expired. GitHub also notes that users on legacy annual plans do not receive access to new models and features.

Model

Multiplier

Claude Haiku 4.5

0.33

Claude Sonnet 4.5

6

Claude Sonnet 4.6

9

Claude Opus 4.5

15

Claude Opus 4.6

27

Claude Opus 4.7

27

Claude Opus 4.8

27

Gemini 2.5 Pro

1

Gemini 3 Flash

0.33

Gemini 3 Pro

6

Gemini 3.1 Pro

6

Gemini 3.5 Flash

14

GPT-4o

0.33

GPT-4o mini

0.33

GPT-5 mini

0.33

GPT-5.1

3

GPT-5.1-Codex

3

GPT-5.1-Codex-Mini

0.33

GPT-5.1-Codex-Max

3

GPT-5.3-Codex

6

GPT-5.4

6

GPT-5.4 mini

6

GPT-5.5

57

Raptor mini

0.33

MAI-Code-1-Flash

0.33

Two details soften and sharpen the picture. Using auto model selection in Copilot Chat, the Copilot CLI, or the Copilot cloud agent earns a 10% discount, so a model with a multiplier of 1 is billed at 0.9. Copilot code review, by contrast, carries a multiplier of 13, so each review deducts thirteen premium requests from the monthly quota.

Source: Model Multipliers for Annual Plans on Request-Based Billing (legacy)


7. Per-Token Pricing Under Usage-Based Billing

For everyone on usage-based billing, which is to say monthly subscribers and organisations, the cost of additional usage is set by per-token rates published by GitHub. GitHub groups models into three categories. Lightweight models such as GPT-5 mini are cheapest, Versatile models such as Claude Haiku 4.5 sit in the middle, and Powerful models such as Claude Opus 4.7 or GPT-5.5 are the most expensive. A short chat using a Lightweight model consumes a fraction of a credit, while an agentic session on a Powerful model consumes far more. All prices below are per one million tokens.

7.1 OpenAI

Model

Category

Input

Cached input

Output

GPT-5 mini

Lightweight

$0.25

$0.025

$2.00

GPT-5.4 nano

Lightweight

$0.20

$0.02

$1.25

GPT-5.4 mini

Lightweight

$0.75

$0.075

$4.50

GPT-5.4

Versatile

$2.50

$0.25

$15.00

GPT-5.3-Codex

Powerful

$1.75

$0.175

$14.00

GPT-5.5

Powerful

$5.00

$0.50

$30.00

GPT-5.4 and GPT-5.5 also offer a long-context tier for inputs above 272,000 tokens, priced higher than the default tier. GPT-5.4 in long context costs $5.00 input, $0.50 cached, and $22.50 output, while GPT-5.5 in long context costs $10.00 input, $1.00 cached, and $45.00 output.

7.2 Anthropic (Claude)

Anthropic models carry an additional cache write cost, charged when the model stores context for reuse.

Model

Category

Input

Cached input

Cache write

Output

Claude Haiku 4.5

Versatile

$1.00

$0.10

$1.25

$5.00

Claude Sonnet 4.6

Versatile

$3.00

$0.30

$3.75

$15.00

Claude Opus 4.7

Powerful

$5.00

$0.50

$6.25

$25.00

Claude Opus 4.8

Powerful

$5.00

$0.50

$6.25

$25.00

Claude Sonnet 4 and Sonnet 4.5 share the Sonnet 4.6 rates, and Claude Opus 4.5 and 4.6 share the Opus 4.7 rates. Claude Fable 5 is listed at $10.00 input, $1.00 cached, $12.50 cache write, and $50.00 output, but GitHub records it as currently unavailable.

7.3 Google, GitHub, and Microsoft

Model

Provider

Category

Input

Cached input

Output

Gemini 3 Flash

Google

Lightweight

$0.50

$0.05

$3.00

Gemini 3.5 Flash

Google

Lightweight

$1.50

$0.15

$9.00

Gemini 2.5 Pro

Google

Powerful

$1.25

$0.125

$10.00

Gemini 3.1 Pro

Google

Powerful

$2.00

$0.20

$12.00

Raptor mini

GitHub

Versatile

$0.25

$0.025

$2.00

MAI-Code-1-Flash

Microsoft

Lightweight

$0.75

$0.075

$4.50

Gemini 3.1 Pro also has a long-context tier above 200,000 tokens, priced at $4.00 input, $0.40 cached, and $18.00 output.

7.4 Working Out the Cost of an Interaction

To estimate the cost of a single interaction, multiply the input tokens by the input rate, the output tokens by the output rate, and any cached tokens by the cached input rate, then add any Anthropic cache write cost. Convert the total from US dollars into AI Credits at $0.01 per credit. This arithmetic explains why frontier models on long agentic sessions dominate a credit bill, because the output rate on a Powerful model can be more than ten times the rate on a Lightweight one, and agentic sessions generate far more output than a single chat.

Source: Models and Pricing for GitHub Copilot


8. What Stays Free

Code completions remain unlimited on all paid GitHub Copilot plans. They are counted separately using GitHub's existing per-completion tracking and are not metered under the new billing system. Next edit suggestions, a separate feature that predicts the change you are likely to make next based on your recent edits, also remain unlimited.

The split between metered and unmetered features is what determines a real bill. The everyday autocomplete that most developers rely on does not consume credits. The billing change affects chat interactions, agentic coding sessions, and code review, which are the features that call models repeatedly and generate large volumes of tokens. A developer who lives mostly in completions will barely touch their credit allowance, while a developer who runs agentic sessions on Powerful models all day can exhaust it quickly.


9. Copilot Code Review Carries Two Costs

Copilot code review is billed in two ways at once. Token consumption is billed in AI Credits, and the agentic infrastructure that powers the review consumes GitHub Actions minutes, charged at the same per-minute rates as other GitHub Actions workflows. A single review therefore generates two separate cost lines rather than one.

There is a further wrinkle that makes code review hard to budget. GitHub selects the review model automatically and does not disclose which model it uses, so the per-review token cost cannot be predicted in advance. The two cost lines also fall on different accounts. AI Credits are charged to the person who requests the review, or to the pull request author when a policy triggers the review automatically, while the Actions minutes are attributed to the repository and from there to the enterprise or cost centre. Teams that turn on automatic code review across many repositories should watch both lines, because the Actions component is easy to overlook when it sits in a separate part of the bill.


10. How Seat Billing Works

10.1 End-of-Cycle Billing

GitHub Copilot charges at the end of each billing cycle based on the number of seats assigned, rather than at the start. You can add or remove seats at any time during the cycle, but a seat you remove is still charged through to the cycle's end. Planning seat changes around the cycle boundary therefore avoids paying for capacity you no longer use.

10.2 Enterprise-Level Consolidation

If your organisation uses GitHub Enterprise Cloud and a user holds Copilot seats in several organisations within the same enterprise, the enterprise is billed only once for that user. One organisation is selected and billed for the seat, and a detailed usage report shows which organisation that is.

Example: Your enterprise has three GitHub organisations, called Development, Platform, and Data. A developer with seats in all three is billed as one seat, not three.

Without Enterprise Cloud, each organisation bills separately, so a user with seats in several standalone organisations is paid for several times. For organisations weighing the wider trade-off between agreement types, our comparison of EA versus CSP and our guide to what Microsoft agreement to choose set out how the purchasing channel changes the commercial picture.


11. What Copilot Enterprise Adds Over Business

The $20 premium for Copilot Enterprise buys more than a larger credit allocation, though that is part of it. Enterprise includes 3,900 credits per user against 1,900 on Business, which is 2,000 additional credits for $20 more per user per month, alongside priority access to new models and features and the ability for an enterprise owner to assign Copilot Enterprise or Copilot Business to individual organisations within the enterprise. Business, by contrast, provides access to the model catalogue with centralised management and policy control for organisation members.

One feature has moved out of the Enterprise-only column. Copilot Spaces, which are curated collections of documentation, wikis, and code that Copilot can search when answering questions, are now available to anyone with a Copilot licence, including Copilot Free. A developer can ask "How do we handle authentication in our API?" and Copilot will answer from the documentation and code patterns gathered in the space, rather than offering generic suggestions. Because Spaces are no longer restricted to Enterprise, they are no longer a reason on their own to choose the higher plan. Questions asked inside a space count as Copilot Chat requests and consume AI Credits in the normal way.

GitHub also offers its own fine-tuned models, such as Raptor mini and Goldeneye, across its paid plans.


12. Prerequisites and Requirements

12.1 GitHub Copilot Business

GitHub Copilot Business is for organisations on a GitHub Free or GitHub Team plan, or for enterprises on GitHub Enterprise Cloud. A GitHub organisation is free to create, and the organisation owner or billing manager assigns seats to users. Note that, as covered in section 3.2, self-serve sign-up for Business on GitHub Free and GitHub Team is paused.

12.2 GitHub Copilot Enterprise

GitHub Copilot Enterprise requires a GitHub Enterprise Cloud subscription with an enterprise account enabled. The enterprise owner controls policy and seat management across all organisations within the enterprise. GitHub Enterprise Cloud is a separate subscription with its own pricing, which adds to the total cost of adopting Copilot Enterprise and belongs in any business case. Copilot is not available for GitHub Enterprise Server.


13. Annual Subscribers and the Transition

Annual Copilot Pro and Pro+ subscribers occupy a transitional position. They remain on request-based billing until their plan expires, with the increased model multipliers from section 6 in effect. GitHub is not renewing annual plans, so at expiry these subscribers move to Copilot Free, with the option to upgrade to a paid monthly plan.

Subscribers who do not want to wait can convert to a monthly paid plan before the annual term ends, and GitHub provides pro-rated credits for the remaining value of the annual plan. There is also a separate cancellation route for individuals who hold a personal plan and are then assigned a Business or Enterprise seat, in which case the personal plan is automatically cancelled with a pro-rated refund.

Alongside the billing change, GitHub tightened usage limits on individual plans in April 2026. It introduced session and weekly token limits, restricted Opus models on Copilot Pro while keeping Opus 4.7 available on Pro+, and began displaying usage limits in VS Code and the Copilot CLI. These limits are separate from credit allowances and act as guardrails on heavy, parallelised usage.


14. Managing Costs

14.1 Monitor AI Credit Usage

GitHub provides usage reports that show how AI Credits are being consumed. Reviewing them monthly identifies users approaching their allowance, the heaviest workloads, and opportunities to adjust model defaults. Because the credit pool is shared, a single heavy user can absorb a disproportionate share, which makes per-user visibility more important under pooled billing than it was under fixed allowances.

14.2 Set Organisation Policies

Administrators can configure which models developers may use. Powerful models such as Claude Opus 4.8 or GPT-5.5 consume far more per interaction than Lightweight or Versatile models, so restricting the most expensive models to specific teams keeps consumption predictable. Encouraging auto model selection also helps, both because it routes simple tasks to cheaper models and because it earns the 10% discount described in section 6. The link between model choice and cost is the same one that drives Copilot economics across Microsoft's wider AI portfolio.

14.3 Right-Size Your Plan

Assigning Business to those who work mostly in code completion, and reserving Enterprise for heavy chat and agent users, lowers the per-user average without starving anyone of capacity.


🖐 Benchmark AI-related licensing and cost exposure. Learn more: Pricing Research and Pricing Metrics.


15. Procurement Checklist

When evaluating GitHub Copilot for your organisation:

  • Clarify which Copilot: Ask the requester whether they need GitHub Copilot for development or Microsoft 365 Copilot for productivity. These are separate products, typically bought under different contracts.

  • Confirm the contract owner: GitHub agreements are separate from Microsoft agreements. Identify who owns the GitHub relationship, because it may not be the team that manages Microsoft licensing.

  • Check your GitHub plan and availability: If you need Copilot Enterprise, confirm you already hold a GitHub Enterprise Cloud subscription, and add its cost to the business case. Confirm current sign-up availability, because self-serve Copilot Business sign-ups for organisations on GitHub Free or GitHub Team remain temporarily paused.

  • Run a pilot with usage tracking: Start with Business on a small team. After one month, pull the usage report to see actual credit consumption per user before committing to a wider rollout.

  • Set budgets before go-live: Configure budgets at enterprise, organisation, cost centre, and user level with hard stops enabled, so overage cannot surprise you in the first billing cycle. Remember there is no automatic fallback to a cheaper model when a budget runs out.

  • Set a model policy from day one: Decide which model categories each team can access. Without a policy, developers default to the most capable and most expensive models.

  • Define seat assignment rules: Decide who can grant seats, what happens when someone leaves, and how mid-cycle additions are approved. Note that you keep paying for a removed seat until the end of the cycle.


16. Enterprise Licensing Considerations

GitHub Copilot sits outside traditional Microsoft licensing agreements, yet it is available through a Microsoft Enterprise Agreement or the Cloud Solution Provider channel, so many organisations now manage it alongside the rest of their AI cost base.

Before an agreement renewal that bundles Copilot, request the GitHub Copilot usage report for your organisation, so you can see which models your developers actually use and how many AI Credits they consume each month before you commit to a figure.


Frequently Asked Questions

How much does GitHub Copilot cost per user?

For organisations, GitHub Copilot Business costs $19 per user per month and GitHub Copilot Enterprise costs $39 per user per month. For individuals, Copilot Pro is $10 per month, Copilot Pro+ is $39 per month, and Copilot Max is $100 per month. Every paid plan includes a monthly GitHub AI Credit allowance, and usage beyond it is billed at per-token rates that vary by model.

What is the difference between GitHub Copilot and Microsoft Copilot?

GitHub Copilot is an AI coding assistant for software developers that runs inside code editors such as VS Code and Visual Studio. Microsoft 365 Copilot is a productivity assistant for knowledge workers that runs inside Word, Excel, Outlook, and Teams. They are separate products with separate pricing and contracts.

What are GitHub Copilot AI Credits?

GitHub AI Credits are the billing unit for GitHub Copilot usage-based billing, in effect since 1 June 2026. Each credit equals $0.01 USD. When you use Copilot Chat, the Copilot CLI, agentic coding, Copilot Spaces, or code review, the interaction consumes tokens that are converted into credits. Code completions and next edit suggestions remain unlimited and do not consume credits.

Is there a free or higher individual plan?

Yes. Copilot Free offers limited access at no charge, and Copilot Student is free for verified students. Above Copilot Pro and Pro+ sits Copilot Max at $100 per month, which includes the highest individual credit allowance at 20,000 credits and priority access to new models. Copilot Pro, Pro+, and Student are open to new sign-ups; Copilot Max is available as an upgrade for users who already hold a Copilot plan.

Does GitHub Copilot work with JetBrains?

Yes. GitHub Copilot integrates with JetBrains IDEs including IntelliJ and PyCharm. It also works with Visual Studio Code, Visual Studio, Eclipse, Xcode, Neovim, and GitHub.com.

Do I need GitHub Enterprise Cloud for Copilot?

Only for Copilot Enterprise. Copilot Business works with organisations on GitHub Free, GitHub Team, or GitHub Enterprise Cloud. Copilot Spaces, which used to require Enterprise, are now available to anyone with a Copilot licence, so they are no longer a reason on their own to choose Enterprise.


Need Help?

If you need help understanding how the new GitHub Copilot billing model affects your agreement and renewal costs, get in touch. We don't sell Microsoft licences or cloud services, so our advice is unbiased.

Table of contents
print
share

Read next

More articles