Codex overage allowance slashed to ~1%; exploit value > disclosure bounty

nptacek · x · 2026-08-25

A user observed that a specific feature in OpenAI's Codex has been virtually eliminated, with its overage allowance now reduced to roughly 1%, matching Claude's limits. The quoted post argues that the value of exploiting a discovered vulnerability is significantly higher than the $1.50 creator revenue gained from exposing it to the public.

Original post →

More from Models

Models channel →