OpenAI Pro $200 is reopening — why does the limit look like half?
At a glance. Pro $200 is reopening, but the new usage accounting can make the allowance look like roughly half of the old Pro $200 when measured in API dollars. Tibo’s framing is that model efficiency gains and lower API costs still let you get more done than a month ago. The 5-hour limit is not coming back, and the extra subscription perks previewed for tomorrow remain unpublished.
Tistory original: donggrri.tistory.com/170
This post restates the 2026-09-29 X note from Tibo (@thsottiaux, OpenAI Codex) for developers and agent users. It does not invent price tables, regional rates, or performance numbers beyond what that source allows.
What reopens tomorrow (Sep 30)?
One-line answer: On 2026-09-30, OpenAI Pro $200 subscriptions open again for new and returning sign-ups.
The headline from Tibo’s Sep 29 post is operational, not speculative: the Pro $200 window that had been closed is scheduled to reopen the next day.
“New and reopen” covers first-time purchases and people who can subscribe again after the window closed. This article is an explainer of that announcement, not a checkout guide or affiliate path.
Exact UI copy and checkout details belong to the official product surface when it goes live. Here we only pin the date and the plan name.
If usage accounting changes, what does the limit feel like?
One-line answer: Under the new accounting, the allowance can look like about half of the old Pro $200 when converted to API dollars.
Keeping the “Pro $200” label does not mean the meter is unchanged. When the way usage is counted shifts, the felt ceiling moves with it. Tibo’s comparison is explicit: roughly half the old Pro $200 in API-dollar terms.
That “half” is not a spreadsheet recalculated from a public rate card. It is the comparison frame stated in the source (old plan vs new accounting).
So a dashboard that shows fewer API dollars can feel like a cut. The next section covers how that can sit next to a claim that you can still do more work than a month ago.
The ~half API-dollar claim is from Tibo’s post. Account, region, and billing period can change what you see — trust your own subscription UI and official docs for the live number.
Why claim you can do more than a month ago?
One-line answer: Because higher model efficiency and lower API costs are framed as flowing into subscription value, so even if the meter looks halved, throughput can still beat last month.
API-dollar headroom and real work completed are different axes. If the model finishes the same job cheaper and faster, the same subscription can run more loops.
Tibo’s logic is that those efficiency and cost improvements are passed into the subscription. That is how “looks like half on the meter” and “more work than a month ago” are presented together.
This post does not invent bench numbers to prove that claim. Validate it against your own workload, where these factors matter more than a single headline ratio:
- how often you push long contexts
- how dense your tool calls are
- how long your agent loops run
What about the 5-hour cap and weekly usage?
One-line answer: The 5-hour limit is not returning, and weekly/usage allotments are framed so you can burn them when you want.
There is no statement that the familiar 5-hour window comes back. The framing is that it does not return.
Weekly and usage pools are described as something you can spend when you choose within the period, not as a force-split schedule that dictates when you may work.
In practice, batch heavy agent runs (large refactors, bulk review) against the weekly/usage remainder, and keep light interactive use flexible. Exact reset times and remaining-quota UI live in the product.
Planning tip: drop workflows that assumed a 5-hour window. Plan big jobs against the weekly/usage remainder instead — that matches the source framing.
How do Sol·Luna at 50% and longer subs fit?
One-line answer: Sol·Luna at 50% sits in the same efficiency and API-gap story; the source does not claim artificial limit inflation.
The source points mention Sol·Luna at 50% price alongside the Pro reopen narrative. Read it as part of the same “efficiency and API economics feed subscription value” thread, not as a separate performance review.
No Sol·Luna bench scores, personal review stats, or long-term discount tables are invented here. Only the stated 50% price mention and the efficiency/API-gap connection appear.
If you are weighing a longer commitment, put reopen timing, the usage-accounting change, and tomorrow’s extra perks (next section) on one checklist, and wait for official terms before locking details.
What extra benefits were previewed for tomorrow?
One-line answer: Extra subscription benefits previewed for tomorrow are said not to reduce usage further; the details are still unpublished.
The preview is that additional subscription benefits land tomorrow and do not further cut usage. Naming those benefits, numbers, or eligibility here would be speculation, so they stay unpublished in this post.
For developers, treat “usage accounting change” and “extra benefits” as same-day siblings: one affects felt limits, the other adds subscription value whose specifics need the official drop.
Follow OpenAI / Codex official announcements and in-product notices only. This post includes no affiliate, reseller, or discount-hub links.
FAQ
Q. Do existing Pro $200 subscribers automatically move to the new accounting tomorrow?
The source centers on new/reopen access and the accounting change. Per-account migration timing and UI copy are for official docs — this post does not assert an automatic switch.
Q. If it is “half in API dollars,” is that an automatic loss?
Not under the source framing. Even if the meter looks halved, efficiency and lower API costs are said to enable more work than a month ago. Validate against your own workload.
Bottom line
Tomorrow (Sep 30) Pro $200 reopens, and the new usage math can look like ~half the old plan in API dollars. The same note argues efficiency and lower API costs still raise what you can get done versus a month ago, without bringing back the 5-hour limit. Treat Sol·Luna 50% as part of that efficiency story, leave tomorrow’s extra perks as unpublished, and follow official OpenAI/Codex announcements only.