Codex tip: 1M context window setting silently kills cache and burns your quota
altryne · x · 2026-09-27
altryne shares a PSA for Codex users: an experimental setting setting the context window to 1M tokens breaks OpenAI's prompt cache and rips through your usage quota. After removing the setting and moving sub-agents to a cheaper model, usage went back to normal. Check your codex.toml for stale experimental settings.
Related event: Codex users warned: 1M context setting destroys cache and burns quota(3 posts)→
More from coding & agent
- Why OpenAI's Mysterious '64H' Might Be About Giving Agents Real-Time Time Budgets — 1a3orn · 2026-09-27
- PHASEONE agent's redacted "64H" hints the real IP is real-time budgets, not token budgets — 1a3orn · 2026-09-27
- Yacine returns to work: frontier models let one person red-team like entire teams used to — yacineMTB · 2026-09-27
- dhh: programmers who can beat a pair of collaborating frontier models round to zero — suchenzang · 2026-09-27
- Claude subscribers can claim $100–$250 in free Claude Code cloud session credits until Oct 7 — Expert_Annual_19 · 2026-09-27
- Agent fixed what its reviewer flagged—then the next review passed the new bug it created — Fit-Ant2238 · 2026-09-27