GitHub Copilot Shifts to Usage Based Billing for High Compute Agentic Workflows

GitHubGitHub

Starting June 1, GitHub Copilot will transition all subscription plans to a usage-based model powered by monthly AI Credits. This shift ends the era of unlimited flat-rate access as autonomous agentic workflows drive inference costs beyond sustainable levels. Users will now pay for actual token consumption once their base credit allotment is exhausted.

GitHub is transitioning all Copilot plans to usage-based billing on June 1, 2026, replacing flat-rate access with GitHub AI Credits. Subscriptions include a monthly credit allotment equal to their price, with extra usage billed by token consumption (the fundamental unit of text processing). Standard autocomplete remains included without consuming credits.
Copilot Pro plan
$10 per month ($10 credits)
Copilot Enterprise plan
$39 per user per month ($39 credits)
Included features
Autocomplete and Next Edit suggestions
Billing metric
Token consumption (input, output, cached)
Effective date
June 1, 2026
Business promotion
$30 to $70 monthly credits (June-August)

This shift addresses the escalating compute demands of GitHub's Fleet Mode sessions. Unlike simple chat, autonomous agents perform multi-step reasoning that requires significantly more inference-time compute (allocating more resources during generation). This mirrors OpenAI's usage-based Codex seats, signaling a move away from unlimited subscriptions for agents.

You can view a preview bill in early May to project costs. Organizations can pool credits across users and set granular budget caps. While GPT-5.5 support remains available, fallback experiences to lower-cost models will be removed. Copilot Pro remains $10 per month, while Enterprise stays at $39 per user.

View the full update on github.blog
GitHub
GitHub
@github
X

Starting June 1st, GitHub Copilot will move to a usage-based billing model as GitHub Copilot supports more agentic and advanced workflows. In early May, you'll see a preview bill experience, giving visibility into projected costs before the transition. 👉 Read more about the upcoming change: https://t.co/4IC9VNHwhk

934retweets2.9klikes
View on X

Still wondering? A few quick answers below.

GitHub AI Credits are the new currency for Copilot usage starting June 1, 2026. Every subscription plan includes a monthly allotment of credits equal to the plan's price. These credits are consumed based on token usage, including input, output, and cached tokens, according to the published API rates for the specific AI models being used.

The transition to usage-based billing officially begins on June 1, 2026, for all GitHub Copilot plans. To help users prepare for the change, GitHub is launching a preview bill experience in early May. This allows individuals and administrators to see their projected costs and usage patterns before the new credit system takes effect.

Standard code completions and Next Edit suggestions remain included in all GitHub Copilot base plans and do not consume AI Credits. These core features are covered by the monthly subscription fee. However, more advanced agentic workflows, long-running coding sessions, and repository-wide iterations will consume credits based on the volume of tokens processed.

Business and Enterprise plans maintain their current monthly seat pricing but now include a matching amount of AI Credits. Organizations can pool these credits across all users to prevent stranded capacity. Administrators gain new budget controls to set spending limits at the enterprise, cost center, or individual user level once the included pool is exhausted.

Once you exhaust your monthly credit allotment, you can no longer use high-compute features unless you or your administrator authorize additional spending at published API rates. GitHub is removing fallback experiences that previously moved users to lower-cost models. Usage is now strictly governed by available credits and the budget controls set by account administrators.

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

GitHub

See all AI news & updates from GitHub →

Share this update