GitHub Freezes Copilot Plans Amid Rising AI Costs
Microsoft-owned GitHub has paused new subscriptions for Copilot Pro, Pro+, and Student plans as of April 20. As a result, the company is responding to rising computing demands from AI coding agents. Moreover, internal signals indicate that usage costs now exceed the limits of the current pricing model.
At the same time, the company is preparing to shift toward token-based billing. Instead of fixed requests, users will likely pay based on actual compute consumption. Consequently, the change reflects a broader adjustment to resource-intensive AI workloads.
Strain from Agentic Workloads
AI-driven coding sessions have become longer and more resource-intensive. Therefore, they now consume significantly more compute than earlier usage patterns. In addition, GitHub has introduced tighter session controls and weekly token caps to manage demand.
The platform has also removed certain advanced AI models from lower-tier plans. Meanwhile, higher-tier users retain limited access, although model availability will continue to change. As a result, the service structure now prioritizes efficiency over unrestricted access.
These adjustments follow earlier disruptions, including paused free trials and corrected usage tracking. Consequently, many users experienced sudden limit exhaustion after system updates.
Transition Toward Usage-Based Pricing
Internal data shows that operating costs for Copilot have increased sharply in recent months. As a result, GitHub is prioritizing a transition to token-based pricing. This approach will align user charges with actual compute usage rather than fixed interaction limits.
Although the timeline remains uncertain, the shift has gained urgency due to growing demand. Therefore, the company is actively developing a more sustainable billing framework.
User Response and Platform Adjustments
GitHub now allows existing subscribers to cancel and request refunds for April usage under specific conditions. At the same time, developers have raised concerns about reduced value under stricter limits.
In response, GitHub has introduced usage tracking tools within its development environments. These tools help users monitor consumption and adjust usage patterns accordingly. Meanwhile, the platform encourages switching to smaller models for less complex tasks.
Overall, the changes signal a transition period as GitHub recalibrates its AI services for long-term sustainability.
