As of June 1, 2026, all GitHub Copilot plans run on usage-based billing. Premium request units are gone. What replaced them is a token-metered currency called GitHub AI Credits: one credit equals one cent, and every model interaction converts into credits based on the input, output, and cached tokens it consumes, charged at each model's published rate.

GitHub's framing is that Copilot outgrew its old pricing. A one-line completion and a multi-hour autonomous run used to cost the same, and once agentic use went mainstream, that flat rate stopped matching the compute behind it. Tying the price to tokens fixes the mismatch.

If your Copilot use is mostly autocomplete, this barely registers. If you drive Copilot as an agent from the terminal, it changes which moves cost money. Here's the practical shape of it.

Requests out, tokens in

Old model: each interaction cost one premium request, scaled by a per-model multiplier, drawn from a monthly request allowance.

New model: each interaction costs whatever its tokens cost on the model you picked. Every paid plan still ships with a monthly pool, now denominated in credits, with the option to set a budget for usage past it. Published figures put the included pool at 1,500 credits for Pro, 7,000 for Pro+, and 20,000 for Max, with pooled per-user allowances on Business and Enterprise.

Worth knowing if you pay yearly: annual Pro and Pro+ subscribers stay on the request-based model until the term ends, and several model multipliers went up for them on June 1. An annual plan doesn't dodge the change. It postpones part of it while making the strong models eat more of the old allowance.

Autocomplete is untouched

Before anyone starts rationing, here's the part that didn't move.

Inline completions and Next Edit Suggestions are still unlimited and still free. If your day is mostly tab-completion in the editor, your costs read identical to May. Nothing to monitor there.

The meter lands on the rest: chat, and especially the agentic runs that open files, plan, run commands, and iterate. That surface is the one that ballooned, and it's the one the new pricing is built around.

The terminal is where it bites

In the CLI you decide which model answers. Until June, that was a quality call and nothing else. Reaching for the strongest model on a trivial task cost nothing you could see.

Now the per-token prices sit behind that choice, and the gap between models is large. One analysis of GitHub's published rates pointed out that output on the priciest listed model runs many times the cheapest, wide enough that grabbing a heavyweight model by reflex is basically a billing bug. Spinning one up to reflow a paragraph isn't just wasteful anymore. It's billable waste.

So for agentic and CLI-heavy work, two ordinary habits become cost decisions: the model you default to, and how long you let a session run. A sprawling all-day agent session now carries a price tied to its length, not just its result.

What "heavy" costs

The included pools feel roomy until you hold them against real agent use. On the GitHub community thread, a Pro+ user reported spending around 360 credits in one ordinary development day and projected a normal month running well past the included 7,000.

That's one workload, not a benchmark, and how near you get depends entirely on how agentic your days are. The signal isn't the number, it's the shape: under the new system, an ordinary day has a credit cost, and for agent-leaning work that cost can rise faster than the monthly pool implies.

New guardrails, and a double charge

GitHub shipped budget controls with the switch, including a hard stop at your limit and a usage view that shows credits draining over the month. There's also a pay-as-you-go path, so you can keep working on your current plan after the included credits run dry instead of being pushed to upgrade.

The wrinkle teams should flag: Copilot code review moved to an agentic architecture, and as of June 1 it consumes GitHub Actions minutes on private repositories on top of AI Credits. A single pull request review can now hit two parts of the bill, the model cost and the runner minutes. Public repos keep their free Actions minutes.

The open question

Cheaper or pricier isn't one answer. For autocomplete-first users it's about flat. For agent-first CLI users the bill now tracks how hard they run sessions and which models they reach for, which can land either way depending on discipline.

Two things to keep an eye on. The price spread rewards matching the model to the task, so a deliberate default plus selective escalation is now a cost strategy, not only a quality one. And the budget hard stop is the cleanest defense against a surprising month, the kind of setting that's better on before you need it than after.

The pattern underneath is the one GitHub says out loud: agentic coding has cloud economics attached now. The terminal feels it first, because the terminal is where the long, autonomous, token-hungry sessions actually run.

---

Sources: GitHub's announcement "GitHub Copilot is moving to usage-based billing" and the June 1 changelog "Updates to GitHub Copilot billing and plans" (github.blog). Included-credit figures from GitHub Docs; the per-day usage figure from GitHub Community discussion #192948; model-rate analysis from published Copilot per-token rates.