Copilot’s code suggestions are about to get more expensive.
Starting June 1, 2026, GitHub is rolling out a revised pricing structure for its AI coding assistant, ditching the simple monthly subscription for a model that blends base credits with a nebulous “flex allotment.” This isn’t just a minor tweak; it’s a fundamental shift that smells suspiciously like corporate calculus trying to squeeze more cash out of developers who’ve come to rely on its… well, assistance.
Developers have apparently been grumbling that the existing usage limits just don’t cut it anymore. Longer AI agent runs? Multi-step coding workflows? More capable — and thus, more resource-intensive — models? GitHub’s heard you. Or, more accurately, they’ve seen the data suggesting they can charge more.
The new lineup features Free, Pro, Pro+, and Max. The Free tier remains limited, but the paid plans are where the real fun (read: expense) begins. Each paid plan now consists of two parts: Base credits, which are directly tied to your subscription price (so $10 buys you $10 in base credits, $39 buys $39, and so on), and the Flex allotment. This Flex part is the kicker – it’s additional usage that may change over time. Sounds like a euphemism for ‘we’ll hike this up when we feel like it.’
Let’s break down the numbers, shall we? The Pro plan, previously $10 flat, now offers $15 total, but $5 of that is the flex allotment. Pro+ jumps from $10 to $39 for $70 total ($39 base, $31 flex). And the new Max plan will set you back $100 for $200 total ($100 base, $100 flex). So, on the Pro plan, you’re paying $5 more for an extra $5 worth of credits, half of which is… variable. Brilliant.
How it works is deceptively simple: your base credits are used first. Then, the flex allotment kicks in. You don’t need to manage anything separately; your dashboard will supposedly show you what’s up. Code completions and next-edit suggestions? Those remain unlimited, because, of course, those are the cheap tricks.
Why Flex Can Change
The explanation for the flex allotment’s variability is pure corporate speak. It’s designed to “adapt as the economics of AI evolve, including model pricing, new models, and improvements in efficiency.” In plain English: as models get cheaper to run or newer, more powerful ones become available, they reserve the right to adjust how much that “flex” is actually worth to you, while your base credits remain stubbornly fixed to the price you pay.
This feels less like an adaptation to developer needs and more like a thinly veiled attempt to decouple pricing from actual, predictable usage. It’s the tech equivalent of buying a lottery ticket where the prize amount is a moving target.
Your base credits, which are 1:1 to your subscription price, will always stay the same.
They say that with a straight face. The problem is, once you’re hooked on Copilot’s convenience, the actual cost of sustained, high-volume use becomes the real story. This flex system introduces an unnerving level of unpredictability into what many developers now consider a core tool. It’s a gamble.
What you’re supposed to do if you’re on a monthly Pro or Pro+ plan? Nothing. It’ll just happen when you migrate to usage-based billing on June 1st. A delightful surprise, I’m sure.
This move echoes some of the less popular shifts in cloud computing pricing over the years, where opaque billing structures and variable costs left many users scratching their heads and their wallets. GitHub is betting that the convenience of Copilot is strong enough to absorb this new complexity and potential cost increase. They might be right, but it’s a risky bet that could alienate a significant chunk of its user base. Developers are used to paying for what they use, but they also like knowing what they’re paying for, especially when it comes to a tool that promises to boost productivity.
This isn’t about making Copilot better in terms of its output; it’s about making it more profitable under the guise of “adapting to developer needs.” The free tier, presumably, will remain the bait, while the real costs lurk in the upgraded plans for those who dare to embrace AI-assisted coding at scale.
Will This Change How Developers Use Copilot?
Potentially. The introduction of a variable “flex allotment” could make developers more conservative with their AI usage. If the cost of extensive coding sessions becomes unpredictable, some might revert to more manual coding, especially for less critical or experimental tasks. This could ironically slow down adoption of advanced AI features if the financial risk outweighs the perceived benefit.
Is GitHub Copilot Still Worth It?
For many, yes. Copilot’s ability to speed up routine coding tasks and suggest solutions is undeniably powerful. However, the new pricing structure demands closer scrutiny. Developers will need to actively monitor their usage and understand the true cost implications of the flex allotment. If your usage falls consistently within the combined base and flex credits without overages, it might still be a good deal. If you anticipate frequently exceeding your allotted flex, the cost could quickly escalate beyond what many developers are willing to pay.
What Does ‘Usage-Based Billing’ Mean for Copilot?
It means that beyond a certain threshold (your base credits plus the flex allotment), you’ll be charged for additional usage. While GitHub states that code completions themselves remain unlimited, other features like chat interactions and agent runs likely consume credits from the flex allotment. The specifics of what constitutes ‘usage’ and its cost beyond the flex limit will be crucial to watch as this model rolls out. It shifts from a predictable subscription to a metered service.
🧬 Related Insights
- Read more: [Voice AI File Manager] Neat Hack, Real Mess?
- Read more: Daily Briefing: April 25, 2026
