GitHub Copilot, the AI-powered coding assistant owned by Microsoft, is about to get pricier for some users.
On Friday, GitHub introduced a new feature called “premium requests,” which imposes rate limits for tasks that use advanced AI models, such as “agentic” coding and multi-file edits. Users can still make unlimited requests with the base model, OpenAI’s GPT-4o, but switching to newer, more capable models like Anthropic’s 3.7 Sonnet will now come with a cap.
Pricing Changes and New Plans
Starting May 5, customers on the Copilot Pro plan ($20/month) will receive 300 premium requests per month. Those on the Copilot Business plan will get 300 requests, while Copilot Enterprise users will receive 1,000 premium requests per month. This change will go into effect between May 12 and May 19, depending on the plan.
For users who exceed their monthly premium requests, they can either purchase additional requests at $0.04 per request or opt for the new Copilot Pro+ plan, which starts at $39 per month. Copilot Pro+ offers 1,500 premium requests per month and access to the best models, including OpenAI’s GPT-4.5.
The Reasoning Behind the Price Increase
The pricing shift, which follows a similar rate increase from AI coding platform Devin, reflects the higher computational demands of more advanced models. While reasoning models like 3.7 Sonnet offer more reliable results due to additional fact-checking, they require more resources to run, contributing to higher costs.
Despite the price increase, Copilot continues to be a profitable venture for Microsoft. CEO Satya Nadella revealed last August that GitHub Copilot accounted for over 40% of GitHub’s revenue growth in 2024 and is already larger than GitHub’s entire business was when Microsoft acquired the platform seven years ago.
What The Author Thinks
While GitHub Copilot’s price hike may be seen as a reflection of rising operational costs, it’s also a reminder of how much AI’s more advanced models truly cost to operate. These models bring tangible value to users, but the additional charges might deter some from adopting them, especially in an increasingly competitive AI space. Companies need to balance affordability and performance, or risk alienating their user base.
