News & Updates, Code Editors, Industry Analysis

GitHub Copilot Drops Flat-Rate Billing, Moves to Token Credits June 1

GitHub announced all Copilot plans will move to AI Credit token billing on June 1, replacing fixed request units. Agentic sessions and chat will now cost based on actual token consumption — a structural shift that has sparked immediate developer backlash.

4 min read
GitHub Copilot Drops Flat-Rate Billing, Moves to Token Credits June 1

Image by GitHub

GitHub Copilot Ends Its Flat-Rate Era: Token-Based Billing Lands June 1

GitHub has confirmed that all Copilot plans will shift from premium request units (PRUs) to a token-based AI Credits model on June 1, 2026, ending the predictable flat-rate pricing structure that made Copilot a competitive staple for individual developers and enterprise teams. The announcement, posted to the GitHub Blog on April 27, landed alongside a temporary pause on new Copilot Pro, Pro+, and Student plan sign-ups that began on April 20.

"Today, a quick chat question and a multi-hour autonomous coding session can cost the user the same amount," wrote GitHub Chief Product Officer Mario Rodriguez. "GitHub has absorbed much of the escalating inference cost behind that usage, but the current premium request model is no longer sustainable."

The change is a direct consequence of Copilot's evolution from a code-completion tool into an agentic platform. Multi-step sessions running across entire repositories, frontier model usage, and the new Copilot cloud agent have all pushed compute costs well above what flat PRU billing was designed to absorb.

What's Changing on June 1

Premium request units will be retired. In their place, every Copilot plan will include a monthly allotment of GitHub AI Credits, where 1 AI Credit = $0.01 USD. Credits are consumed based on token usage — input, output, and cached tokens — priced at the listed API rate for each model.

Monthly subscription prices are not changing:

PlanMonthly PriceIncluded AI Credits
Copilot Pro$10/month$10 in credits
Copilot Pro+$39/month$39 in credits
Copilot Business$19/user/month$19 in credits
Copilot Enterprise$39/user/month$39 in credits

Two features remain unlimited on all paid plans and will not consume AI Credits: code completions and Next Edit Suggestions. Everything else — Copilot Chat, agent mode, Copilot cloud agent, Copilot Spaces, Spark, and third-party coding agents — will draw from the credit pool.

Copilot code review gets a double billing hit starting June 1: it will consume both AI Credits and GitHub Actions minutes at standard per-minute rates.

The Developer Impact

The practical effect varies sharply by workflow. Developers who primarily rely on inline completions will see little change — that feature stays unlimited. The friction lands on anyone using Copilot Chat heavily, running agentic coding sessions, or regularly reviewing pull requests with Copilot.

GitHub's own documentation is candid: "A long coding agent session using a frontier model across multiple files will cost more, because it's doing more work." For developers running agent mode against large codebases with expensive models like Claude Opus 4.7, the credit pool could deplete faster than expected under the $39 Pro+ ceiling.

The model multiplier question is where annual plan subscribers face the sharpest hit. GitHub is raising PRU multipliers for expensive models effective June 1. Claude Opus 4.7 — currently at a 7.5x PRU multiplier in Pro+ — will jump to a 27x multiplier under the revised annual-plan schedule. Annual subscribers keep their PRUs until the plan expires, but the effective cost of frontier model usage rises significantly before they migrate.

Community Reaction

Developer reaction in GitHub's own FAQ discussion thread has been swift and pointed. Visual Studio Magazine headlined the response: "Devs Sound Off on Usage-Based Copilot Pricing Change: 'You Will Get Less, but Pay the Same Price.'"

Representative comments from the thread include:

  • "This just wiped GitHub's value moat — why should I stay?"
  • "This is removing the one real advantage GHCP had over Claude Code et al."
  • "Request-based billing is user oriented... Token-based billing is not user oriented."
  • "I would much rather you jacked up the price per premium request by 30–50x than switch to token based pricing where agents and so forth can nuke your supply of credits in very little time."

GitHub pre-loaded its FAQ with its own question and answer: "This just wiped GitHub's value moat — why should I stay?" GitHub's response: "We believe GitHub Copilot remains the best value and experience for agentic coding. Usage-based billing aligns cost more closely to actual usage and value."

Transition Details

To ease the migration, GitHub is providing promotional bonus credits for Business and Enterprise customers from June through August: Business users receive an extra $30 in credits per user per month, and Enterprise users receive an extra $70. Pooled credit sharing across an organization will also be available, so lighter users' unused credits offset heavier users within the same team.

Monthly plan subscribers will migrate automatically on June 1 with no action required. Annual subscribers have two options: switch to monthly plans on June 1 (GitHub will provide prorated credits for the remaining annual term), or ride out the annual plan under the revised multiplier schedule until it expires — after which the plan reverts to Copilot Free.

GitHub is also building a billing preview tool, expected in early May, that lets developers upload their usage CSV and see projected costs under the new model before the switch takes effect. Developers dissatisfied with the current restrictions can cancel and receive a refund through May 20, 2026.

What's Unconfirmed

GitHub has not disclosed how many Copilot subscribers are on annual versus monthly plans, or what proportion of current users hit the kinds of agentic workloads that will most affect their credit consumption. The company has not said whether Pro or Pro+ credit limits will be adjusted after the transition period ends. Whether the new billing model successfully reduces wait times and usage throttling — the stated justification for the change — will only become clear after June 1.

Share:

Other Latest News

Vercel Caps Hobby Plan Deployments at 30 Days Starting Today
News & Updates, Deployment

Vercel Caps Hobby Plan Deployments at 30 Days Starting Today

Vercel is hard-capping Hobby plan deployment retention at 30 days starting today, automatically purging older builds. Here's what survives the cull and what developers need to do right now.

Apr 29, 2026
Claude API and Claude.ai Suffer Multiple Outages on April 28
News & Updates, AI Assistants, Industry Analysis

Claude API and Claude.ai Suffer Multiple Outages on April 28

An active outage hitting both the Anthropic API and Claude.ai login paths on April 28 marks the third Claude incident in a single day, raising fresh questions about reliability for Claude Code and API-dependent workflows.

Apr 29, 2026
OpenAI Ships Symphony: Codex Agents Now Run Your Linear Board
News & Updates, AI Agents, API Tools

OpenAI Ships Symphony: Codex Agents Now Run Your Linear Board

OpenAI open-sourced Symphony today — an orchestration spec that maps every open Linear issue to a dedicated Codex agent workspace, removing the per-session supervision bottleneck. Some internal teams reported 500% more landed pull requests in three weeks.

Apr 28, 2026
China Blocks Meta's $2B Manus Acquisition, Orders Deal Unwound
News & Updates, AI Agents, Industry Analysis

China Blocks Meta's $2B Manus Acquisition, Orders Deal Unwound

China's NDRC has blocked Meta's $2 billion takeover of Manus AI and ordered the deal unwound, citing AI tech export concerns — leaving developers facing uncertainty over the platform's future.

Apr 28, 2026
OpenAI Ends Azure Exclusivity: Codex and GPT API Go Multi-Cloud
News & Updates, API Tools, Industry Analysis

OpenAI Ends Azure Exclusivity: Codex and GPT API Go Multi-Cloud

OpenAI and Microsoft jointly announced today that OpenAI can now serve all its products — including Codex and the GPT API — across any cloud provider, ending years of Azure-first exclusivity in a structural shift for every developer building on the OpenAI stack.

Apr 28, 2026
DeepSeek Cuts V4-Pro Prices 75%, Slashes Cache Costs Tenfold
News & Updates, Industry Analysis, API Tools

DeepSeek Cuts V4-Pro Prices 75%, Slashes Cache Costs Tenfold

DeepSeek announced a 75% limited-time discount on V4-Pro plus a permanent tenfold reduction to cache hit costs across its entire API — reigniting the pricing war that China's AI labs keep winning.

Apr 27, 2026
← Scroll for more →