
In April, GitHub announced a shift from request-based billing to usage-based pricing for Copilot. The new system grants paid subscribers a monthly allotment of AI 'credits,' with one credit equal to $0.01 of usage. The $10/month Pro plan includes 1,500 credits ($15 worth); the $39/month Pro+ plan includes 7,000 credits ($70 worth); and the $100/month Copilot Max plan includes 20,000 credits ($200 worth).
Under the old system, GitHub absorbed much of the cost difference between simple queries and multi-hour coding sessions. Now, credit consumption depends on input and output tokens used and the underlying LLM (large language model) chosen—OpenAI's GPT-5.4 nano costs $1.25 per million output tokens via Copilot, while GPT-5.5 frontier model costs $30 per million output tokens.
Many users report extreme sticker shock: some burned through a month's quota in a single day, one user spent 840 credits in limited testing on the first day, and another used 21 percent of their monthly Pro allotment in one day. Users sharing cost estimates from GitHub's own tool showed their previous monthly usage would have cost thousands of dollars under the new pricing. Some Copilot users are threatening to cancel or switch to other AI coding services.
Summaries like this, in your inbox every morning.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Across 24 counted runs of /code-review on a 41-line shopping-cart file with six planted bugs, --max-findings 1…

The tutorial, based on official Claude Code documentation and checked as of October 18, 2026, says failed fixe…

A developer writing on Zenn says he stopped dividing work to keep Claude Code's context small, letting Sonnet…

In a demo on a fictional "Asterlane" page with three service cards, HtmlDrag selected only the middle card and…

A Claude Code user split recurring explanations into knowledge notes, Skills, and CLAUDE.md, and said session-…

Docker published an analysis of major AI coding agent incidents from October 2024 to February 2026, including…
