In a surprising move that could affect numerous developers, GitHub Copilot, the AI-powered coding assistant developed by Microsoft-owned GitHub, is set to introduce new premium pricing limits for its AI models. As the tech landscape evolves, cost considerations are coming into sharper focus for users of this popular coding tool.

On Friday, GitHub announced the implementation of a “premium requests” feature, which places caps on the number of uses for advanced AI models beyond the standard option. While subscribers can still enjoy unlimited access to the foundational model, OpenAI's GPT-4o, any requests to harness the more advanced models – like Anthropic’s 3.7 Sonnet – will now be significantly limited.

For users subscribed to the Copilot Pro plan, which costs $20 per month, the new system will allocate 300 premium requests starting May 5. Business and Enterprise tier users will benefit from an increased limit of 300 and 1,000 premium requests, respectively, with their rollout scheduled between May 12 and May 19.

For those requiring more than the allotted premium requests, additional access can be purchased at a rate of $0.04 per request. Alternatively, users may opt for the newly launched Copilot Pro+ plan, priced at $39 per month, granting 1,500 premium requests along with access to the top-tier AI models, including the latest OpenAI's GPT-4.5.

This pricing adjustment mirrors similar moves by competing platforms, like Devin, which recently raised rates for certain users, indicating a significant trend as the operational costs for advanced AI models continue to rise. Highly capable models, such as 3.7 Sonnet, necessitate more rigorous fact-checking, resulting in increased computational demands and higher costs.

However, it's important to note that despite the added costs, GitHub Copilot remains a robust source of revenue for the tech giant. Microsoft CEO Satya Nadella revealed last August that Copilot contributed more than 40% to GitHub's revenue growth in 2024 and has now outstripped GitHub's overall revenue at the time of its acquisition nearly seven years ago.

As the demand for AI assistance in coding persists, developers are left wondering if the value provided by these enhanced models will justify the added expense. As we move further into the realm of AI in software development, the question remains: How will this shift impact the everyday user of GitHub Copilot? Stay tuned for more updates as we uncover the evolving landscape of coding assistance technology.