OpenAI Scraps the Pro $200 Cut, Doubles Pro 100 and Pro 500 Limits

At 16:56 UTC on October 29, 2026, Thibault "Tibo" Sottiaux posted day 25 of his 28-day update. The opening line says usage limits for Pro 100, Pro 200, Pro 500 and Business Premium double as of today. The rest of the post says the Pro 200 limits will not be reduced, so Work and Codex usage there stays where it is.

The OpenAI help center spelled that out more precisely than the tweet did. Pro 100 and Pro 500 now get twice their previous Work and Codex allowance. What Pro 200 loses is the reduction announced in the September 29 email: accounts that already had the higher allowance keep it, and accounts on the lower allowance are brought up to match. Plus is listed as unchanged.

Read only the first line of the tweet and it's easy to conclude that every Pro tier got another 2x. What he actually reversed was the earlier expiry cut: after the September 30 reopening, new Pro $200 subscribers dropped from 20x to 10x while existing users stayed at 20x. The full background is in OpenAI reopens Pro $200 and adds a $500 tier.

How the tiers actually differ on the official pages

Here is how the help center and the ChatGPT release notes (the October 29, 2026 entry) describe Work and Codex.

Tier	Help center / release notes	Notes
Pro 100	Work and Codex usage at 2x the previous level	Weekly GPT-6 Pro messages in Chat are not listed as doubled
Pro 500	Also 2x the previous level	Ultrafast still not offered below Pro 500
Pro 200, previously higher allowance	The September 29 reduction is canceled; allowance holds	No second multiplication on top of it
Pro 200, previously lower allowance	Lower allowance doubles and is raised to the higher level	No "40x" appears anywhere on the page
Newly purchased Pro 200	Gets that same unified allowance, still $200/month	Canceling and resubscribing won't switch you to a different set
Business Premium	Release notes say usage increases; Tibo says it doubles	Help center gives no per-product multiple
Plus	Unchanged	Plus is not part of this round of increases

Existing subscriptions apply automatically; there is no need to wait for the next billing cycle. The help center states that current subscribers don't need to cancel and resubscribe. Cancel and come back later, and you get the same allowance as every other Pro 200 account.

The same page also sets out three boundaries:

The weekly GPT-6 Pro message cap in Chat is counted separately from Work and Codex allowance. Pro 100 and Pro 500 keep their current weekly cap. On Pro 200, the lower group moves up to the higher cap, and accounts already at the higher cap stay as they are.
A bigger allowance does not mean twice as many tasks get finished. The model, context, reasoning, tool calls and speed mode all change consumption.
Topping up a Pro 200 account does not change the tier. It is still Pro 200, and Ultrafast remains exclusive to Pro 500.

Where the tweet and the help center diverge

Tibo's original post reads:

Day 25/ We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today. We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage.

The source is his own account, status/2115850175868727621.

Tibo's original post on X

Three separate things are packed into one post.

"Doubling all usage limits" for the tiers listed, naming Pro 100, 200, 500 and Business Premium.
"We heard your feedback, so we won't reduce Pro 200" — a rollback of an already announced reduction, with Work and Codex described as staying the same.
Extended dot limits continue, counted separately from default usage. No new end date is given.

Read the first two sentences together and it sounds like Pro 200 got another 2x on top. The help center splits them apart: the doubling lands on Pro 100, Pro 500, and the previously lower group inside Pro 200. For the previously higher Pro 200 accounts, the page says the allowance is maintained. Multiplying by two again never appears.

Business Premium is described with different precision in the two official texts. The release notes' first sentence counts both Pro and Business Premium under "increases Work and Codex usage," but when the next sentence mentions 2x, it names only Pro 100 and Pro 500. Tibo's post says flatly that Business Premium doubles. Both texts come from OpenAI. Where the help center gives no multiple, we treat it as not published rather than filling the number in from the tweet.

Later that same day, several reposts recorded another line from him: Graduating from The Reset Company to The Token Company. That line reads as a statement of position. The post neither promises that resets are over nor describes a new allowance formula.

How this relates to the September 29 halving

On September 29, Tibo said the new Pro $200, converted at API list prices, was worth roughly half the dollar allowance of the old tier, and he wrote the relative multiples as Plus = 1x, Pro 100 = 5x, Pro 200 = 10x. Existing Pro 200 accounts kept 20x for the time being and received extra credits. As of October 29, the help center now says the reduction announced in that September 29 email is canceled.

That single sentence is the only point where the two announcements line up. The multiples below are the September table combined with this round's "2x." The October 29 help center did not reissue a multiple table, so each item carries a "if":

If Pro 100 was still at 5x before the cut, then "2x the previous allowance" puts it at roughly 10x again.
If newly opened Pro 200 accounts were at 10x, doubling them lands on the higher allowance that existing users already hold. 40x does not appear on the page.
If existing users are holding the original 20x, they stay at 20x. 40x is not written there either.
Pro 500 was 25x relative to Plus in September. The help center only says Work and Codex become 2x the previous level, and 50x is never written. Until OpenAI publishes a new table, that number shouldn't be treated as the current price.

The size, delivery and expiry of September's one-time credits are not mentioned in this help center update. The cut was canceled, but the rules for those credits were not updated as a result.

What this update did not touch

Plus usage is unchanged. The 5-hour window does not appear anywhere in the October 29 update. Whether Pro will get that limit back, or whether Plus loses it, is not addressed on the page.

Ultrafast remains a Pro 500 feature. The same page also states that for GPT-6 Astra and GPT-6.1 Sol, Ultrafast first draws on plan allowance at 8x the rate of the selected model's standard mode, and once that runs out it charges purchased credits at 6x the standard credit rate. The wording immediately after adds that these are consumption rates and do not guarantee faster output. On Pro 100 or Pro 200, buying credits is explicitly not a way to unlock Ultrafast.

Dots are still counted separately, and the deadline is still "for now." Codex Reset Watch checked this against the official dots documentation and added that the distinction already existed in those docs — day 25 didn't change it. Chatting with a dot uses the dot's own allowance; tasks it dispatches inside Work or Codex still count against those limits.

The release notes carried one other separate item the same day: eligible ChatGPT Pro and Business subscribers can activate promotional API credits, recorded against their linked API organization and kept separate from ChatGPT usage credits. That runs on its own track from the subscription allowance increase, and the page gives no dollar figure here either.

The help center frames all of this as a plan allowance adjustment that existing users receive automatically. Nowhere does the page describe it as a reset that puts usage back to 100%.

What the community is saying

What follows is public discussion from people online.

Linux.do: which Pro 200 tier actually doubled?

3093927 was posted at 00:55 UTC on October 30. When this article was written the thread had reached floor 32, with roughly 1.6k views and 81 likes. An earlier thread a few minutes before it, 3093393, had only 6 replies.

3S_Real_Man asked whether the "$200 won't be reduced" line means doubling on top of the existing 20x, or dropping to 10x and then climbing back to 20x. Long123456 read it the way the help center does: the plan to cut 20x down to 10x on October 30 was called off. Later replies compressed the beneficiaries into a single line: the $100 tier gets a real doubling, and the higher end of the $200 tier simply avoids a cut.

More replies were spent working out who got left out.

Plus was repeatedly called the odd one out. Some said they would stop renewing next month, some read it as a nudge from Plus toward Pro, and others noted that Team at least still gets bonus allowance.
Existing 20x accounts were described as getting essentially nothing. youer corrected that quickly: the ones who weren't cut recently and still have credits to claim are the old 20x accounts; the ones who were cut for a month with no credits are the new Pro 200.
The distrust came down to three lines: "they're just handing back what should have been ours," "they decide the limits, so it'll drift back down," and "they're secretly billing at double behind the scenes." The last one came with no bills or logs attached.

Others simply said "good, keep it coming." Skepticism outnumbered gratitude.

The three data points below are all from individual accounts, so the sample is the poster:

Niruri reported 21% remaining the day before, then 61% on /status that day, and confirmed in a reply that they are on 20x. If the absolute amount consumed stayed the same and only the cap doubled, a 79% used share would become roughly 39.5% used, leaving about 60.5% — close to 61%. That doesn't line up with the help center's "higher Pro 200 accounts are unchanged." They didn't say which line item /status referred to, and one report isn't enough to overturn the help center's wording.
Another user claiming 20x said the number they see is now around 1300, versus around 2500 before. No unit was given. Replies offered two guesses: that Plus's baseline shrank too, dragging the absolute 20x allowance down with it, or that Astra's launch already took one cut, this update only canceled the second, leaving them near 1500 with no path back to 2500. Both are guesses from the thread, and nobody posted before-and-after usage logs.
One Pro 500 user said Ultrafast's speed dropped from roughly 8x to roughly 2x and suspected compute had been redirected to the lower tiers. The help center's 8x refers to the rate at which Ultrafast burns plan allowance, and the page states that this does not guarantee faster output. So far this is the only report of a slowdown.

Developer forum: higher-allowance Pro 200 users see nothing added

A thread on the OpenAI developer community puts the verdict in its title: Day 25 wasn't a 2x improvement for existing Pro 200 users. The author acknowledges that lower-allowance accounts and the other Pro tiers did get more, then writes that $200 subscribers who already had the higher allowance got nothing extra. The line is: A canceled downgrade is not an upgrade.

They point back to the October 4 public commitment: for the following 28 days, each day would either bring a noticeable improvement for most Codex and Work users or a full reset for everyone. The commitment is reposted in this tracking thread. In their reading, counting "we decided not to make existing subscriptions worse" as the day 25 deliverable means no substantive update was shipped, and the commitment should trigger a reset.

Someone in the same thread agreed and added: at minimum, give the $200 tier a banked reset it can use later. Another short reply said they had been about to subscribe to Claude that day; calling off the cut is good, but don't sell it as an upgrade for heavy users.

Check your own subscription

The only way to know how much your absolute token allowance moved is to look at your own usage page.

Pro 200 subscribed after September 30, on the lower allowance. The help center says the lower allowance doubles and is raised to the higher level, with the price still $200. This is the group where the page actually uses the word "doubling."

Pro 200 that was already on the higher allowance before the halving announcement. The stated outcome is that the reduction is canceled and the allowance stays.

Pro 100 and Pro 500. Work and Codex are at twice the previous level, per the help center. For weekly GPT-6 Pro messages in Chat, the page says these two tiers are unchanged. Pro 500's speed tier also did not become generally available through this increase.

Plus. The help center says unchanged.

At every tier, OpenAI states the same thing: you don't need to cancel and resubscribe to get the new allowance. Subscription allowance will keep shifting with models and settings. It was rewritten once in September and again now, so treating it as stable long-term infrastructure is a bad bet. Pricing can be redrawn inside a week — the same problem shows up in DeepSeek V4's pricing and in the Claude Opus 4.7 feedback.

Still unstated

"2x the previous level" for Pro 100 and Pro 500 comes with no absolute figure and no new multiple against Plus. The September table of 5x / 10x / 20x / 25x is available for comparison, but those numbers are not on the new price sheet yet.
Business Premium doubles in the tweet and "increases" in the release notes.
The "for now" on extended dot limits has no date attached.
Whether September's credits still ship under the original email is not addressed in this update.

FAQ

Did Pro 200 double too?

Per the help center: Pro 200 accounts on the lower allowance double and are aligned with the higher allowance. Accounts already on the higher allowance are described as unchanged. Tibo's "all Pro 100/200/500" opening is broader than that.

Did Plus get an increase?

No. The help center states that Plus usage is unchanged.

Do I need to cancel and resubscribe?

No. The page states that existing subscriptions apply automatically. Cancel and come back, and you get the same unified allowance. No hidden higher tier is described.

Is Ultrafast slower now?

The help center doesn't say that. Its 8x figure is the rate at which Ultrafast consumes plan allowance, and it explicitly says that this does not guarantee faster output.

Does this count as an improvement in the 28 days?

Tibo labeled the day "Day 25." Under his October 4 rule, this day shipped an improvement, and no reset went out. Some in the developer community argue that higher-allowance Pro 200 and Plus saw no capacity increase that day, that canceling a downgrade is not an upgrade, and that the commitment therefore calls for a reset.

Bottom line

The October 29 update, as fixed in the help center, comes down to three things.

Pro 100 and Pro 500 get twice their previous Work and Codex usage. Weekly GPT-6 Pro messages in Chat are not part of that sentence.
Pro 200 cancels the reduction announced on September 29. Higher-allowance accounts are unchanged, lower-allowance accounts are leveled up, and the monthly price stays $200.
Plus is unchanged. Extended dot limits are still counted separately, with no new end date. Ultrafast remains Pro 500 only.

The tweet's first line, some media headlines, and the "20x becomes 40x / 25x becomes 50x" arithmetic all go further than those three points. The community split lands in exactly that gap: lower tiers see allowance restored, higher tiers see only that nothing was taken away, and Plus sees the ladder pulled wider apart.