GitHub Copilot moves to usage-based billing from June 1: credits replace premium request units, Pro plan receives $10 monthly AI Credits

GitHub has announced a significant change to GitHub Copilot’s billing model: starting June 1, 2026, the company is introducing usage-based billing through a system of AI Credits, replacing the previous premium request unit system. The change affects all paid plans, but leaves the most popular feature — code completions — untouched.

The Core Change: Credits Instead of Request Units

GitHub writes in its announcement: “Starting June 1, your Copilot usage will consume GitHub AI Credits.” Instead of a system in which users have a monthly quota of “premium” requests, the new model works through credits consumed based on actual token usage — including input, output, and cached tokens — at the published API rates of each individual model.

In practice, this means expensive models (e.g., the largest reasoning models) consume credits faster, while cheaper models allow more interactions within the same monthly quota. Users gain transparency that did not exist before — no longer asking “how many premium requests do I have left,” but rather “how many dollars of credits do I have left.”

Plan Pricing and Monthly Quotas (Unchanged)

Plan prices remain the same, and each plan receives a monthly AI Credits quota nominally matching the price:

Plan	Price	Monthly credits
Copilot Pro	$10/mo	$10
Copilot Pro+	$39/mo	$39
Copilot Business	$19/user/mo	$19
Copilot Enterprise	$39/user/mo	$39

A free tier exists, but GitHub does not specify a credit quota for it in the available announcement.

What Remains Unlimited

GitHub explicitly preserves what is Copilot’s most widely used feature:

• “Code completions and Next Edit suggestions remain included in all plans and do not consume AI Credits.”

This means the average developer primarily using inline suggestions will not notice a difference in costs.

What Consumes Credits

Credits begin to be consumed when users move to more interactive modalities:

• chat interactions with Copilot,
• autonomous coding sessions (agent loops),
• code review — which additionally consumes GitHub Actions minutes (double billing in that case).

This billing architecture aligns with the industry trend that inference using reasoning is significantly more expensive than simple completions, so costs are shifted to the users who actually consume it.

Enterprise-Specific Changes

Large organizations receive several important additions:

• promotional included usage through August — a period in which teams can test the new system without direct costs,
• pooled credits across the entire organization — credits do not have to be consumed individually,
• new budget controls at the enterprise, cost center, and individual user levels.

This is significant for FinOps teams that previously had limited visibility into what Copilot actually costs per project or department.

Transition for Annual Subscribers

GitHub confirms: “Annual subscribers retain current pricing until expiration, then convert to free tier or paid monthly plans with prorated credits.” So, if you have paid for an annual subscription, you do not need to do anything until the contract expires — current terms remain in effect.

Practical Implications for Teams

Three things worth doing in the next five weeks, before June 1:

• measure current monthly consumption of chat interactions and agent sessions per developer — this is a signal of how much credit usage will be under the new model;
• revise internal guidelines — when is it smart to use chat, when inline completion, when a local model for simple tasks;
• prepare FinOps integration — enterprise budgets and alerts by cost center will be available and should be configured before credits actually start flowing.

This change fits into a broader industry trend (OpenAI, Anthropic, Google already charge per token in their APIs), but GitHub had until now maintained fixed monthly pricing for all plans. Moving to usage-based billing brings transparency benefits, but places the burden of cost tracking directly on teams.