GitHub Copilot: Token Billing Divided

Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisis ton rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
The era of fixed pricing for Github Copilot from Microsoft is coming to an end, replaced by a token-based billing model. This change, effective from June 1, could lead to a significant increase in costs for users, particularly for small businesses and freelancers. Large companies may still afford it, but smaller entities will need to reassess their monthly budgets.
Users will now be charged based on the number of tokens consumed during their work, rather than a fixed rate based on requests. Developers, surprised by this new pricing, have expressed their dissatisfaction on platforms like Reddit and X. One user shared that their monthly cost would rise from $29 to $750, stating: "This new pricing model is simply too expensive. I will cancel my subscription. At this price, it is no longer cost-effective or practically useful." Another saw their fees jump from $50 to $3,000.
Some users believe these increases are exaggerated, but others point out that excessive token consumption is often due to inefficient use, referred to as vibe-coding. "The big difference between some of us who work all day and barely have extra costs, and those screenshots. I find it hard to believe that there are differences in complexity in the workload," wrote one user.
The previous economic model of Copilot raises questions. One user asked: "How much money was Copilot losing?" Some wonder how much the company was losing by subsidizing the heavy use of its tool. While critics blame users for their excessive consumption, others argue that Microsoft encouraged this practice. "To everyone blaming… those who actually used the system the way Microsoft designed it (and even encouraged to use it this way), honestly, the only one responsible here is Microsoft. Microsoft provided this billing method and continued to make it easier and easier to consume a large number of tokens on single premium requests that could run for hours or even days, while generating dozens or even hundreds of sub-agents," wrote a user.
TechCrunch reached out to Microsoft for a comment, but received no response before publication.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.