GitHub Copilot adopts usage‑based pricing; developers burn credits in a day

GitHub replaced its $10‑per‑user‑month Copilot plan with a token‑credit system on June 1. Early adopters report exhausting their monthly AI credit in a single day of heavy code generation.

sources[Ars Technica][GitHub Blog]

GitHub rolled out a usage‑based pricing model for Copilot on June 1, 2026, ending the $10‑per‑user‑month subscription that had been in place since 2022 [GitHub Blog]. Under the new plan each user receives a monthly AI credit; once the credit is exhausted, token usage is billed at $0.02 per thousand tokens for the first million tokens and $0.015 thereafter [GitHub Blog].

The announcement defines three tiers: a free tier with 10 k tokens per month, a paid tier granting 100 k tokens for $20, and an enterprise tier with customizable caps. Existing subscribers were automatically moved to the paid tier unless they opted out. The structure mirrors the pay‑as‑you‑go models used by cloud AI providers such as OpenAI.

The shift has three immediate effects. First, budgeting becomes granular: teams can monitor token consumption in real time and align spend with actual feature usage. Second, credit exhaustion is immediate—several developers reported burning their entire monthly credit in a single day of intensive code generation, forcing them to pause Copilot or incur extra charges [Ars Technica]. Third, competitive pressure rises as rivals like Tabnine and Claude market “predictable cost” plans, positioning flat‑rate pricing as an alternative to GitHub’s token model.

Analysis Developers accustomed to a flat monthly bill now face an extra layer of cost management. Early feedback shows many are adjusting workflows or exploring other assistants, giving competitors a lever to attract cost‑sensitive teams.

Reader poll

Which pricing model do you prefer for AI coding assistants?