OpenAI Repeatedly Resets Codex and ChatGPT Work Limits After GPT-5.6 Sol Launch

With the release of GPT-5.6 Sol, OpenAI reset the usage limits for ChatGPT Work and Codex multiple times within 24 hours to encourage users to try more challenging tasks. However, the rollout faced issues; the team acknowledged problems such as high-compute modes consuming limits too quickly, and the frequent resets sparked user discussions regarding quota mechanics.

Official Response and Adjustments

The OpenAI team reviewed the initial feedback and admitted a few missteps in the new update: the highest compute mode was too easy to trigger and consumed too much quota. To address this, they reset the limits and promised another adjustment later that day. Additionally, to clear up user confusion that Codex might be phased out, OpenAI clarified that focusing on ChatGPT Work did not mean abandoning Codex, emphasizing that Codex is here to stay.

User Feedback and Controversies

The frequent limit resets drew mixed reactions from users. @0xKaiser noted their quota was reset three times consecutively—jumping from 20% back to 100%—speculating that this was either due to extra consumption from the new update or dynamic background policy adjustments. Meanwhile, @FHSS97 discussed strategies for account upgrades, pointing out that utilizing a reset to restore usage to 100% before upgrading to a Pro plan offers better value. @alexcovo_eth viewed the weekly reset mechanism with expiration dates as a practical improvement that solves the pain points of quota management.

2026-07-10 ~ 2026-07-11 · 12 related posts

4 near-duplicate retellings: Dimillian · soumitrashukla9 · JeremyNguyenPhD · jxnlco