
Developers are expressing growing frustration over GitHub Copilot's new token-based billing system, effective June 1. Affected developers report price jumps from around $29 to $750 per month. Amid Microsoft Build 2026 in San Francisco, many are left questioning the future of AI coding tools.
The recent billing changes under the new "AI Credits" system charge users for chat and coding sessions based on model rates. Some developers are pointing out that costs for a single agentic coding session can chew through $30โ$40 in credits, far exceeding the price of prior flat-rate models.
"A short chat question can cost the user just as much as an autonomous coding session lasting several hours," remarked GitHub's Chief Product Officer.
Critics are voicing concerns that this model disrupts the creative coding process. One developer mentioned, "Copilot is dozens of different things. Itโs also an extension for Visual Studio for agentic workflows."
Some developers have faced staggering increases in monthly bills. Reports mention increases from about $50 to $3,000, prompting questions about GitHubโs ability to support smaller teams and independent developers effectively.
The potential for companies to build their own models has entered discussions as the financial implications become clear. Some believe that Microsoft may secretly benefit from this shift, hinting, "this might actually force companies to negotiate enterprise deals, which is probably what Microsoft wants anyway."
Today's Microsoft Build 2026 event is expected to reveal the company's new MAI coding model, aimed at being competitively priced on Azure. Observers are keen to see if this model can offer real value.
Interestingly, the MAI model may be benchmarked against select competitors like Anthropic's Haiku, raising further scrutiny about Microsoft's market strategy in AI.
๐ธ Monthly costs for coding tools are skyrocketing, some rising to $750 or more.
โ ๏ธ Developers lament the change in the billing model, fearing it limits coding exploration.
๐ "First, they use your code to train the LLM then they sell it to you," voiced another concerned developer.
As the pricing fallout unfolds, some analysts suggest that about 60% of smaller teams may actively seek alternatives if Microsoftโs MAI model doesn't deliver. High costs could lead to increased interest in open-source solutions or custom tools developed in-house.
The current scenario mirrors the early 2000s music industry's struggle with steep fees. Just as artists sought new paths, developers today may also embrace innovative self-built systems that reduce reliance on expensive corporate offerings.
As Microsoft and GitHub navigate this crisis, the landscape of AI coding tools may be on the verge of significant transformation.