GitHub Copilot gets more expensive: how changes in fees will affect developers
By ForGeeks As of May 5, GitHub Copilot, an AI assistant for writing code, will become more expensive for those who actively use advanced neural networks like Anthropic Sonnet 3.7 or GPT-4.5. Microsoft is introducing a “premium request” system that limits access to those models, leaving only the basic GPT-4o unlimited.
New rules: who pays how much
Users of the Copilot Pro subscription ($20/month) will be able to make up to 300 premium requests per month. For enterprise customers, the limits are higher, with Copilot Enterprise offering 1,000 requests. Those who exceed the quota will be able to purchase requests or upgrade to the new Copilot Pro+ plan for $39/month. This includes 1,500 premium queries and priority access to top models, including GPT-4.5.
Why is GitHub tightening policies?
The main reason is the rising cost of computing. Models like Anthropic Sonnet 3.7 spend more resources on code analysis and error checking, making them more accurate but more expensive to run. Competitor Devin recently made a similar move, raising prices for some users.
A similar move was made recently by competitor Devin, raising prices for some users.
Despite this, Copilot remains profitable. According to Microsoft CEO Satya Nadella, the service accounted for 40% of GitHub’s revenue growth in 2024 and overtook the entire platform in terms of revenue when it was acquired in 2018.
Who will be affected by the changes?”
At risk are teams that use Copilot for complex tasks, they will either have to plan more carefully to use premium models or migrate to Copilot Pro+. Those who work with the basic GPT-4o will not notice the change.
Developer Tips
Experts recommend starting by assessing your needs. If you rarely go beyond code autocompletion, the current rate will suffice. For those who actively use advanced features, it’s useful to track premium request usage statistics in your personal cabinet. For large teams, it may be better to get Copilot Pro+ at once to avoid overpayments.
The market for AI development tools is getting more expensive. Companies are shifting costs to users, but improving models in parallel. This could widen the gap between startups with limited budgets and corporations willing to pay for premium features. For now, GitHub has no plans to cancel the innovations, but community pressure could affect the policy.
