copilot's new token billing has developers doing math at 2am
github completed its switch to usage-based AI credits on june 1 — and the backlash was immediate, with some developers reporting bills jumping from $29 to $750. the metering of vibe coding is officially here.
on june 1, github completed copilot's move to token-based AI credits: every suggestion, chat completion, and code review now draws against a monthly budget instead of living under a flat subscription. the reaction was about what you'd expect — fast, loud, and full of screenshots, including reports of monthly costs jumping from $29 to $750 for heavy agent users.
the trend, not the outlier
if this feels familiar, it's because you watched cursor live through the same movie in 2025 — flat plans replaced by credit pools, confusion, apology, refunds (our review covers the aftermath). the entire industry is converging on the same uncomfortable truth: agentic coding consumes wildly different amounts of compute per user, and flat pricing can't survive contact with someone running background agents all day.
the $29→$750 stories are real but they're also the tail: that's what happens when yesterday's “unlimited” habits meet today's meter. the median developer will land somewhere much less dramatic. still — the days of not knowing what a prompt costs are ending everywhere.
what to actually do
our advice is the same regardless of which tool you run: know your usage shape before you pick a plan. if you mostly autocomplete, metered billing is probably cheaper for you. if you lean on long agent runs, a bundled subscription (claude's max tiers, cursor's ultra) is the hedge. and either way: check the meter weekly for the first month. nobody should learn their usage pattern from an invoice.