How to see GitHub Copilot cost per user
By Jiri Sifalda
To see GitHub Copilot cost per user, add two numbers for each person: the seat price and the AI credits that person consumed. GitHub gives you both, but in three different places, and none of them shows the sum.
This guide shows where each number lives, who can read it, and how to combine them into one cost per user.
Why the seat price is not the cost per user
GitHub lists Copilot Business at $19 per granted seat per month and Copilot Enterprise at $39 per granted seat per month (GitHub Copilot plans). Under usage-based billing, each seat also adds AI credits to a shared pool: 1,900 per user per month on Business and 3,900 on Enterprise, at 1 AI credit = $0.01 USD (usage-based billing for organizations and enterprises).
The same page lists the details that make a per-user number hard to read:
- Credits are pooled at the billing entity level. One heavy user can spend credits that a light user did not use, so the invoice does not show who used what.
- Additional usage is enabled by default on Business and Enterprise. When the pool is empty, usage continues and is charged, unless an administrator disables the AI credits paid usage policy.
- Inline code completions and next edit suggestions are not billed in AI credits. Chat and agent work consume credits. Completions do not.
- The pool resets at 00:00:00 UTC on the first day of each month, and unused credits do not carry over.
So the cost of one user is their seat price plus the value of the credits they consumed. The seat price is the part you already know.
Option 1: the AI usage report
The AI usage report is the most detailed per-user source. GitHub describes it as "a detailed per-user breakdown of AI credits consumed for a maximum period of 31 days" (billing reports reference).
To get it, open the AI usage page in your billing settings, click Get usage report, specify the report details, and click Email me the report. GitHub sends a download link to your primary email address, and the link expires after 24 hours (viewing usage of metered products). The same page states that organization owners cannot view user-level data directly, and that they must download a usage report instead.
The report sums usage by date, model and username. The fields that matter for a cost per user are:
username: the user associated with the usage.model: the AI model used.quantity: the amount used.gross_amount: the cost before discounts.discount_amount: the part covered by discounts, including your included usage.net_amount: the billable amount. Gross minus discount equals net.input,output,cache_readandcache_write: the tokens behind the credits, per model. GitHub added these in August 2026 (GitHub changelog).
Use gross_amount to compare users. The included pool covers part of the usage, so net_amount tells you what the organization paid, not what each person consumed.
Option 2: the Copilot usage metrics API
If you want a daily pull instead of an emailed file, use the Copilot usage metrics API. Since June 19, 2026, the user-level reports include an ai_credits_used field for each user (GitHub changelog). The endpoints are in the Copilot usage metrics REST reference.
Know its limits before you build on it:
- It is one total per user. It does not split credits by model or feature.
- GitHub calls it a metrics signal for analyzing consumption, not a billed total. Reconcile your sums against the invoice.
- Only enterprise administrators and organization owners with access to Copilot usage metrics can read it.
- Data for one day is available within two full UTC days after that day closes (About Copilot usage metrics).
Option 3: the seat list, for the fixed part
The seat price is fixed, but you still need to know who holds a seat. GET /orgs/{org}/copilot/billing/seats lists each seat with its assignee, plan_type, created_at, last_activity_at and pending_cancellation_date (Copilot user management REST reference). Only organization owners can call it, and classic tokens need the manage_billing:copilot or read:org scope.
last_activity_at reflects IDE activity only when the user has telemetry enabled in the IDE. An empty value does not prove that a seat is unused.
What each user can see
Since July 2026, Copilot Business and Enterprise users can see how many AI credits they used this billing cycle on their Copilot usage page in settings. If an administrator set a budget for them, they see their use against that budget (GitHub changelog). This helps each developer manage their own use. It does not give a manager a view across the team.
Put the cost per user together
- List the seats. Pull the seat list and note each user's plan, so you know their seat price: $19 or $39 per month.
- Pull the credits. Use the AI usage report for a monthly review, or
ai_credits_usedfor a daily view. - Convert credits to dollars. Multiply credits by $0.01, or take
gross_amountfrom the report. - Add the two. Seat price plus credit value is the cost per user for the month.
- Mark the delay. The metrics API lags by up to two UTC days. Show the last complete day, so nobody reads a gap as low use.
- Reconcile. Compare the sum of all users against the invoice, and record the difference.
Cap the cost, not only measure it
GitHub also supports user-level budgets, which cap how many AI credits one user can consume in a billing cycle. A budget is a hard stop: credit features are blocked, completions keep working, and the seat stays assigned until the next cycle or until an administrator raises the budget (budgets for usage-based billing). Measure first, so that you set each budget from real use and not from a guess.
Copilot is rarely the only AI cost
Most teams also pay for a second tool, such as Claude Code, which reports usage in a different place and unit. The per-developer AI spend tracker guide shows how to join both into one row per developer.
BurntAI is built to do this for you: every developer's Copilot cost next to their other AI spend, against a budget, with a month-end forecast. See pricing to get on the list.