GitHub Copilot to Switch to Usage-Based Billing on 1st June as AI Pricing Shifts Industry-Wide

A Major Pricing Overhaul

Microsoft-owned GitHub has announced that all GitHub Copilot plans will transition to usage-based billing on 1st June 2026, replacing flat per-seat fees with a system of 'GitHub AI Credits' tied to token consumption. Under the new model, one AI credit equals $0.01, with credits consumed based on the tokens each interaction uses. GitHub framed the change as an important step toward a sustainable Copilot business, noting that usage-based billing better aligns pricing with actual usage.

An Industry-Wide Pivot

The Copilot announcement is the latest signal that the software industry's pricing playbook is undergoing its most notable transformation in years. By the end of 2025, the number of enterprise software firms shifting away from flat per-user fees had more than doubled from a year earlier. HubSpot moved its AI Customer Agent pricing on 14 April to an outcome-based model, charging $0.50 per resolved conversation. Salesforce has introduced 'agentic work units', Workday launched 'Flex Credits' as a prepaid digital wallet, and ServiceNow has adopted a hybrid model with a guaranteed floor and flexible consumption metrics. Adobe charges credits for AI-generated images through Firefly.

Why the Shift Is Happening

The economics of AI are forcing the change. Running large language models and AI agents requires expensive GPU compute, and flat per-seat pricing does not account for the wide variation in how much AI any given customer actually consumes. As of 2026, 74% of software suppliers had adopted some form of usage-based pricing model, according to a Stripe analysis. SaaS prices increased by an average of 8 to 12% in 2025, with what industry observers call an 'AI tax' driving much of the increase. Goldman Sachs analysts, after meeting with roughly 40 software and internet companies, observed that companies are increasingly positioning AI workflows as selling a unit of labour or productivity, opening the door to larger deal sizes.

Market Reaction

Application software stocks rose on Monday, with the Nasdaq Composite reaching all-time highs following a recovery from an April dip. The rally reflects growing investor confidence that leading software companies can monetise AI rather than be disrupted by it. For enterprise customers, the trade-off is real: software spending could become less predictable, but vendors gain the ability to charge based on value delivered rather than headcount served.