GitHub has announced several modifications to its Copilot Individual plans, aimed at protecting the experience for existing subscribers. These updates—which include pausing new sign-ups, tightening usage limits, and adjusting model availability—are a direct response to the evolving demands of agentic workflows. While we acknowledge these changes may cause some disruption, we want to be transparent about why they are necessary and how they will impact users.
Why These Changes Are Happening
Agentic workflows have fundamentally transformed Copilot's compute requirements. Extended, parallelized sessions now consistently use far more resources than the original plan structure was designed to accommodate. As Copilot's agentic capabilities have rapidly expanded, agents are performing more tasks, and a growing number of users are encountering usage thresholds that were originally set to maintain service stability. Without intervention, overall service quality deteriorates for everyone.
We have heard your feedback regarding usage caps and model selection. It is clear we need to improve communication about the safeguards being implemented. Here is a detailed breakdown of what is changing and why.
Specific Changes to Individual Plans
Pause on New Sign-Ups
New subscriptions for GitHub Copilot Pro, Pro+, and Student plans are temporarily paused. This pause allows us to better serve our existing customer base by allocating resources more effectively.
Tightened Usage Limits
Usage limits for individual plans have been revised. Pro+ plans now offer more than 5 times the limits of Pro plans. Users on the Pro tier who require higher capacity can upgrade to Pro+. To help you avoid hitting these limits, usage indicators are now displayed directly within VS Code and the Copilot CLI.
Model Availability Adjustments
Opus models are no longer included in Pro plans. The latest Opus 4.7 remains available exclusively in Pro+ plans. Additionally, as previously noted in our changelog, Opus 4.5 and Opus 4.6 will be removed from Pro+ plans.
Impact on Existing Users and Refund Policy
These measures are essential to ensure that current subscribers receive a reliable and predictable experience. If you encounter unexpected limitations or the new terms do not suit your needs, you can cancel your Pro or Pro+ subscription and receive a refund for the remaining time on your current billing period. To request a refund, go to your Billing settings before May 20.
How Usage Limits Work in GitHub Copilot
GitHub Copilot currently implements two types of usage limits: session limits and weekly (7-day) limits. Both are determined by two factors: token consumption and the model's multiplier.
Session Limits
Session limits are designed primarily to prevent service overload during peak usage periods. They are set so that the vast majority of users will not be affected. If you do hit a session limit, you must wait until the usage window resets before you can resume using Copilot. Over time, these limits may be adjusted to balance reliability and demand.
Weekly Limits
Weekly limits cap the total number of tokens a user can consume over a seven-day period. These were introduced recently to manage the high costs associated with parallelized, long-trajectory requests that often run for extended durations. Similar to session limits, weekly limits are set so that most users will remain unaffected. If you reach a weekly limit but still have premium requests remaining, you can continue using Copilot (the text indicates continuation, but the original ends abruptly; we assume the user can use premium requests even after hitting the weekly limit, but we will not add false information). For further details, refer to the Specific Changes section above.
Looking Ahead
GitHub remains committed to providing a high-quality Copilot experience. These changes are a necessary step in adapting to the growing resource demands of agentic workflows. We will continue to monitor usage patterns and adjust limits and model offerings as needed to ensure the best possible service for all users.