GitHub moves Copilot to AI Credits billing on June 1

What changed for Copilot users

GitHub is changing Copilot from a pricing model that mostly hid heavy usage into one that meters it directly. In the official post on X, the company said GitHub Copilot will move to usage-based billing on June 1 and that users will see a preview billing experience in early May. That is a material shift because modern Copilot usage is no longer just short autocomplete bursts or quick chat prompts. It increasingly includes long agentic coding sessions that consume far more inference.

“Starting June 1st, GitHub Copilot will move to a usage-based billing model as GitHub Copilot supports more agentic and advanced workflows.”

The linked GitHub blog post spells out the mechanism. Premium request units will be replaced by GitHub AI Credits, consumed according to token usage across input, output, and cached tokens. Base plan prices stay the same, but the included usage becomes explicit: Copilot Pro keeps its $10 monthly price with $10 of credits, Pro+ stays at $39 with $39 of credits, Business stays at $19 per user, and Enterprise stays at $39 per user. GitHub also says Business and Enterprise customers will receive promotional included usage for June through August, and unused credits can be pooled across an organization instead of being stranded per seat.

Why this is bigger than a billing tweak

The github account usually saves this kind of post for product changes with broad platform impact, not small experiments. The company is effectively admitting that Copilot has become an agentic compute product, not just an IDE add-on. The blog says a quick question and a multi-hour autonomous coding session can no longer be treated as the same economic event. It also notes that code completions and Next Edit suggestions stay included, while fallback experiences disappear once included credits are exhausted. A separate GitHub changelog adds another cost signal: on private repositories, Copilot code review will also start consuming GitHub Actions minutes on June 1.

What to watch next is how teams adjust budgets once real usage becomes visible. Some organizations may embrace heavier agent workflows once they can pool credits and set per-user budgets; others may discover that review loops, long-running coding agents, and premium models cost more than they assumed. Either way, GitHub has moved Copilot pricing much closer to the actual economics of inference. Source: GitHub source tweet · GitHub billing post · GitHub code review billing changelog