Dev reports ChatGPT/Codex burning through usage allowance 20% faster

RileyRalmuto · x · 2026-08-22

A developer who has been building the same app with Codex for months reports that token usage is visibly burning down faster than normal — at least a 20% delta for identical work. Either the model is consuming more tokens, or the allowance itself was quietly reduced. The author later noted it seemed slightly better but couldn't confirm.

Related event: Users Report Abnormal Token Consumption in OpenAI Codex(3 posts)→

Original post →

More from Models

Models channel →