If you have been bumping into Codex’s usage ceiling lately, there is good news. OpenAI has temporarily lifted the five-hour usage limit across its Codex and ChatGPT Work plans, reset everyone’s usage, and pushed efficiency improvements to its flagship GPT-5.6 Sol model. The changes were shared on Sunday, July 12, 2026, by Thomas “Tibo” Sottiaux, who leads the effort at OpenAI.
“The last 48 hours of Codex and ChatGPT Work have been intense,” Sottiaux wrote, describing a surge in demand following the recent GPT-5.6 Sol launch. His post outlined three changes:
Morning. The last 48 hours of Codex and ChatGPT Work have been intense! Three important updates: – Temporarily removing the 5 hour usage limit restriction for all Plus, Business and Pro plans – Rolling out changes that will make GPT 5.6 Sol more efficient across the board…
— Tibo (@thsottiaux) View post on X
- The 5-hour limit is gone (for now). The rolling five-hour window that previously interrupted long sessions has been temporarily removed for all Plus, Pro, and Business subscribers, so you can keep working without hitting that cap.
- GPT-5.6 Sol is more efficient. OpenAI is rolling out changes that make Sol consume less per query “across the board,” meaning your usage stretches further and the same allowance gets you more done.
- Usage got reset. On top of the above, OpenAI issued a one-time usage reset to give subscribers extra headroom on the flagship model right away.
It is worth reading the fine print: these are temporary measures, and OpenAI has not published an end date. The reset also does not make Sol truly unlimited — it simply hands you a fresh, larger runway. This follows a pattern from the days around the Sol launch, when OpenAI reset rate limits multiple times so people could take ambitious tasks for a spin without immediately running dry. Community developers noted the timing lined up with Codex crossing roughly six million users.
OpenAI just temporarily removed the 5 hour usage limit from Codex. Usage with GPT 5.6 Sol feels unlimited now.
— BridgeMind (@bridgemindai) View post on X
The reaction has been telling. The relief was real — some users said Sol suddenly “feels unlimited,” and the move landed particularly well because a common gripe in the preceding days was that Sol chewed through the five-hour allowance alarmingly fast, sometimes with little visibility into why. Not everyone read it as pure generosity, though; a fair share of developers saw the resets as a competitive response in an escalating price and usage-limit war between the big AI labs, rather than a gift. Either way, the practical outcome for users is the same: more room to build, at least for now.
For anyone building with these tools, the practical takeaway is simple: right now is an unusually generous window to lean on GPT-5.6 Sol for heavier work. That is especially relevant if you have been experimenting with running Codex and GPT-5.6 Sol inside Claude Code — with the cap relaxed and Sol running leaner, delegating big reviews or refactors to it costs you less of your allowance than it did a week ago. Just keep in mind the relaxed limits are on borrowed time, so it is a good moment to test what these models can really do before the usual restrictions return.
Leave a Reply