Press Esc to close

OpenAI is reopening $200 Pro, but the usage math is changing

Abstract illustration of a teal-to-coral usage gauge whose needle has been turned back from a dashed earlier position, beside a purple calibration dial circled by arrows, suggesting usage being recalibrated

OpenAI's $200 Pro plan is about to take new subscribers again. There's a catch, though, and OpenAI is getting it out of the way before DevDay.

Just before midnight on Monday in San Francisco, Tibo, who works on Codex and ChatGPT at OpenAI, posted on X that Pro $200 would reopen to new subscribers "tomorrow," along with a change to how its usage is calculated:

About an hour later, he cleared up the timing: "That was yesterday, today is DevDay." So the reopening should arrive with today's DevDay announcements. The keynote starts at 10 a.m. PT, which is 10:30 p.m. IST tonight. We previewed what might show up yesterday.

Now the catch. In Tibo's words, "if you do the math, it will net out at half the dollar in API spend compared to the old Pro $200 plan." Put simply, priced at API rates, a Pro subscription will cover half as much as it used to. That's his framing, and OpenAI hasn't published the new formula.

He still argues you'll get more done than a Pro subscriber did a month ago. His case has four parts.

First, OpenAI is "committing to not reintroducing the 5h limit," so you can use your weekly allowance whenever you want. That rolling five-hour cap sat on top of the weekly limit in Codex and ChatGPT Work. OpenAI temporarily lifted it for Plus, Business and Pro on 12 July, then brought it back for Plus on 25 August, keeping it off for Pro "for the upcoming months." Now, for Pro $200 at least, that reprieve becomes a promise.

Second, he says the subscription will get you more work, at higher quality, over time, as models get more efficient and OpenAI passes the savings on as API price cuts. His example is GPT-6 Sol and GPT-6 Luna. OpenAI's announcement on 22 September cut API prices for both by 50% compared with GPT-5.6 promotional pricing. Sol now costs $2/$10 per million input/output tokens, down from $4/$20, and Luna costs $0.10/$0.50, down from $0.20/$1.20.

Third, he says OpenAI doesn't want to "artificially inflate the API list prices" to make subscriptions look generous. Over time, he expects API prices to fall low enough that most people can simply buy usage as needed.

Fourth, OpenAI is adding things to the subscription that "won't draw on the usage." He wouldn't say what they are.

Why was Pro closed in the first place? OpenAI paused new Pro $200 sign-ups and upgrades on 10 September, a week after GPT-6 Astra launched, because the tier put "the most strain on our systems." As of this afternoon, OpenAI's help center still describes the pause. Pro $100 (5x Plus usage, versus 20x on Pro $200) stayed on sale, as did Plus at $20 a month.

To be clear, everything about the reopening comes from Tibo's posts. It isn't on an official OpenAI page yet.

Credit where it's due: flagging a cut before the keynote beats burying it in a help-center page. But for heavy Codex users, "half" is the number that will stick. Cheaper models help, but GPT-6 Astra, the model that caused the capacity crunch, wasn't part of last week's price cut. We'll update this once OpenAI shares the details.

Comments