Copilot just 9x'd Sonnet and 27x'd Opus and teams have no idea

2026-04-29 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Cloud Computing & IT Infrastructure · Depth: Intermediate, medium

Summary

GitHub Copilot has significantly increased the multipliers for its premium AI models, Claude Opus 4.6 and Sonnet 4.6, effective June 1. Opus's multiplier jumped from 3x to 27x, and Sonnet's from 1x to 9x, meaning these models now consume "premium requests" much faster. This change signals the end of a heavily subsidized pricing model, driven by escalating compute costs for AI providers like Anthropic and Microsoft. Many corporate users, provisioned with Copilot as a benefit, have been indiscriminately using the most advanced models like Opus without visibility into model-level consumption. This shift is a precursor to GitHub's move to full usage-based billing starting June 1, which will directly charge companies based on actual model usage, potentially leading to substantial budget overruns for unprepared organizations.

Key takeaway

For CTOs and VP of Engineering overseeing AI tool adoption, your teams must immediately audit and adjust AI model usage patterns before June 1. The shift to usage-based billing means that continued indiscriminate use of high-cost models like Claude Opus will lead to significant, unexpected budget expenditures. Prioritize implementing model governance and routing strategies to match task complexity with the most cost-effective model, or face substantial financial penalties.

Key insights

AI model providers are ending subsidized pricing, shifting to usage-based billing reflecting true compute costs.

Principles

• Compute costs dictate AI model pricing.
• Indiscriminate use of frontier models is unsustainable.

Method

Providers are implementing higher multipliers and transitioning to usage-based billing to align consumption with actual compute expenses, moving away from flat-rate subscriptions.

In practice

• Implement smart model routing for AI tasks.
• Monitor model-level consumption for cost control.

Topics

• GitHub Copilot
• AI Model Pricing
• Compute Costs
• Anthropic Opus
• Usage-Based Billing

Best for: CTO, VP of Engineering/Data, Machine Learning Engineer, AI Engineer, Software Engineer, Director of AI/ML

Related on AIssential

• GitHub Copilot AI token charges to go up 10×–100× — AI chatbot services are moving to usage-based pricing, revealing the high, previously subsidized operational costs.
• Microsoft Cancels Internal Anthropic Licenses As Shift To Token-Based AI Billing Blows Up Annual Budgets In Months — Token-based AI billing models are proving prohibitively expensive for large-scale enterprise adoption, leading to rapid budget depletion.
• The All-You-Can-Eat AI Buffet Is Over (And Your Codebase Is the Leftovers) — The era of heavily subsidized AI compute and "all-you-can-eat" token usage has ended, shifting costs directly to enterprises.
• Is this the dawn of the Tokenpocalypse? — AI's high operational costs are forcing a shift to usage-based pricing, challenging profitability and business models.
• Here's What Indie Devs Need to Do Before Claude's Pricing Split — Anthropic is ending programmatic Claude usage subsidies, shifting to metered billing for automated tasks.
• What a $26K AI Bill Really Reveals — The true cost of AI is not token spend, but the hidden business dependencies it creates.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.