June 2026 Copilot billing migration
GitHub Copilot Billing Preview Calculator
Paste the AIC total from GitHub's billing preview or usage report, choose your Copilot plan, and estimate included credits, extra spend, pooled-credit pressure, and power-user budget guardrails before usage-based billing takes effect.
What this preview calculator is for
GitHub is moving Copilot from request-based billing to usage-based billing on June 1, 2026. Under the new model, Copilot interactions consume GitHub AI Credits based on input tokens, output tokens, cached tokens, and the model used. Code completions and next edit suggestions remain outside AI Credits for paid plans, but chat, CLI, cloud agent, Spaces, Spark, code review, and third-party coding-agent usage can draw from the credit allowance.
GitHub's official preparation flow gives organization and enterprise admins a preview usage report. The report includes the fields that matter most for this page: aic_quantity, which is the number of AI Credits consumed, and aic_gross_amount, which is the estimated dollar amount under usage-based billing. GitHub also provides a browser-based billing preview tool that keeps the uploaded data in the browser. Use that official preview first, then use this page to turn the top-line AIC number into decisions about budgets, additional usage, and power-user policy.
How to read the result
The included pool is calculated from the plan and the number of seats. Individual plans use the published monthly AI Credit totals for Pro, Pro+, and Max. Organization plans use the per-user pool for Business or Enterprise. GitHub says Business and Enterprise credits are pooled at the billing entity level, so a team with uneven usage can benefit from light users offsetting heavy users. That is useful, but it can also hide a governance problem: one or two agent-heavy users may be responsible for most of the credits.
The overage output is the part of the AIC total that exceeds the included pool. Since GitHub AI Credits draw down additional budgets at a fixed rate of one cent per credit, the dollar overage is easy to estimate. The calculator compares that overage with the additional budget you enter. A budget that is technically enough may still be a bad rollout policy if most of the overage comes from unclear experiments, automatic reviews on private repositories, or a few users choosing expensive models for routine work.
| Input | Where to get it | Decision it supports |
|---|---|---|
| Total AIC | GitHub preview tool or usage report aic_quantity. |
Whether the included pool is enough for current behavior. |
| Seats or users | Copilot license count for the billing entity. | Monthly included pool and average credits per user. |
| Additional budget | Your planned organization or enterprise budget. | Whether usage continues or blocks after included credits are exhausted. |
| Power users | Top users in the GitHub CSV or premium analytics. | User-level budget, enablement, and model-routing rules. |
Migration checklist before June 1
- Download or view the GitHub usage preview for the most recent representative period.
- Separate normal editor completions from AI Credit consuming features.
- Find the top users, top models, and top feature categories by AIC.
- Decide whether additional usage is allowed, blocked, or allowed only for named users.
- Set budgets before teams discover that long agent tasks can consume credits much faster than old premium request counts suggested.
Common budget patterns
A small engineering team may choose a soft organization budget with alerts only. That works when users understand model cost and are mostly doing interactive chat. A larger team should usually add user-level budgets for power users and a policy for expensive model choices. A platform team running automated code review needs an extra check because GitHub says Copilot code review is billed in AI Credits and the agentic infrastructure also consumes GitHub Actions minutes on GitHub-hosted runners starting June 1, 2026.
The safest policy is not "disable everything." It is controlled routing. Keep cheaper models as the default, reserve frontier or Opus-class models for work that truly needs them, require explicit approval for background agents, and measure accepted outcomes. A high-credit task can be worth it if it saves real engineering time. A low-value summary repeated across many repositories can become expensive waste.
FAQ
Why does this page ask for AIC instead of premium requests?
Premium request counts do not map cleanly to the new billing model. GitHub's preview data already converts real usage into AI Credits, so AIC is the cleaner planning input.
What if I do not have GitHub's preview CSV yet?
Use the separate Copilot AI Credits calculator to estimate token-based usage by model, seats, and tasks. Return to this page after you have a real AIC total from GitHub.
Should I use the promotional Business or Enterprise allowance?
Use it only if you are an existing customer in the June 1 to September 1, 2026 promotional window. Treat the promotional number as a measurement window, not a permanent budget baseline.
Sources
- GitHub Docs: Preparing your organization for usage-based billing
- GitHub Docs: Usage-based billing for organizations and enterprises
- GitHub Docs: Usage-based billing for individuals
- GitHub Docs: Models and pricing for GitHub Copilot
- GitHub Copilot AI Credits Calculator
AI Code Limits is independent and is not affiliated with GitHub, Microsoft, Copilot, OpenAI, Anthropic, or Google.