OpenAI may pause new Pro signups as GPT-6 Astra demand surges

By Dennis Singleton · Markets Writer · September 9, 2026, 9:48 AM ET

Key points

OpenAI may pause new Pro subscriptions, its Codex lead says
He calls demand for GPT-6 Astra unprecedented
Pro starts at $100 a month and carries the top Astra access
Astra's rollout left paying users waiting

OpenAI may stop selling new Pro subscriptions if demand for its new GPT-6 Astra model keeps climbing, according to the person who leads Codex at the company. Thibault Sottiaux wrote on X at 1:34 a.m. Eastern on Wednesday that demand is unprecedented and that OpenAI's pulling every lever it has to keep up. Pro is the top consumer tier and starts at $100 a month.

The post had drawn 2.9 million views and 2,000 replies by Wednesday morning.

Demand for Astra is really unprecedented. We're pulling all the levers possible to sustain the demand, but I've not seen anything like it until now and we went through very steep growth before. Priority will always be to keep excellent service for existing users, but we might have to pause new Pro subscriptions for a bit if this continues.
— Tibo (@thsottiaux) September 9, 2026

What Pro buys

Pro sits above the $20 Plus tier and the $8 Go tier, and OpenAI's pricing page lists it at "from $100 a month" for five times or 20 times the usage of Plus. It's the only consumer plan that includes what the company calls Pro reasoning powered by GPT-6 Astra, along with the largest allowances for Codex tasks, deep research, memory and image generation. A pause on new Pro subscriptions would close the top rung of that ladder to new buyers while leaving existing subscribers in place.

Sottiaux didn't say when a pause might start, how long it would last, or whether Plus and Go would be affected. Sottiaux described a possible pause, not a decision to stop accepting subscriptions.

The rollout had already left paying users waiting

OpenAI began releasing GPT-6 Astra on September 3, starting with enterprise customers in its gated Daybreak program before moving to ChatGPT subscribers. Paying users waited days, and on the evening of September 3 Sottiaux said the company would give subscribers who couldn't yet reach Astra one banked usage reset for each day of the wait, Implicator.ai reported. Chief executive Sam Altman said of the rollout, "When we screw up, we try to make it right."

An OpenAI spokesperson told the same outlet the staged release was meant to buy time to add compute capacity, because Astra is "a very large model." OpenAI buys that capacity from Microsoft (MSFT) Azure and Oracle (ORCL), among others, and runs it on Nvidia (NVDA) chips. Altman has cited about $1.4 trillion in infrastructure commitments over eight years.

Frequently asked questions

Is OpenAI pausing new Pro subscriptions?

Not yet. Thibault Sottiaux, who leads Codex at OpenAI, wrote on X at 1:34 a.m. Eastern on September 9, 2026 that the company might have to pause new Pro subscriptions for a bit if demand for GPT-6 Astra continues. He said the priority will always be to keep excellent service for existing users. He gave no start date or duration, and described a possible pause rather than a decision to stop accepting subscriptions.

How much does ChatGPT Pro cost?

OpenAI's pricing page lists Pro at from $100 a month, buying five times or 20 times the usage of the $20 Plus tier. Pro sits above Plus and the $8 Go tier and is the only consumer plan that includes what OpenAI calls Pro reasoning powered by GPT-6 Astra, along with the largest allowances for Codex tasks, deep research, memory and image generation.

Why is OpenAI short of capacity for GPT-6 Astra?

An OpenAI spokesperson told Implicator.ai that the staged release starting September 3, 2026 was meant to buy time to add compute capacity because Astra is a very large model. Paying subscribers waited days for access, and Sottiaux said the company would grant one banked usage reset for each day of the wait. Chief executive Sam Altman said of the rollout that when the company screws up, it tries to make it right.

Which public companies supply OpenAI's computing power?

OpenAI buys computing capacity from Microsoft (MSFT) Azure and Oracle (ORCL), among others, and runs its models on Nvidia (NVDA) chips. Chief executive Sam Altman has cited about $1.4 trillion in computing commitments over eight years. Sottiaux described a possible pause on new Pro subscriptions, not a decision to stop accepting them.

More on MSFT and NVDA

News

Microsoft (MSFT) wants developers to run AI models on their own PCs instead of paying for cloud tokens

Project Zenith is a preconfigured Windows 11 experience for PCs with 64 GB of unified memory, arriving first on AMD's Ryzen AI Halo. Microsoft says it helps reduce reliance on metered cloud tokens. Independent tests of the same chip show what runs fast and what doesn't.

Sep 7, 2026

News

AMD's AI chip gets an 11-fold performance boost from software updates, SemiAnalysis says

The benchmark firm that called AMD's long-context software 'horrible' on August 24 says the MI355X is now 11 times faster on the same MiniMax M3 agentic test, from software alone. The written update with the numbers hasn't been published yet.

Sep 7, 2026

News

Intel (INTC) heads to the AI Infra Summit September 15 with Crescent Island questions unanswered

Intel CEO Lip-Bu Tan takes the main stage at the AI Infra Summit in Santa Clara on Tuesday, September 15, with three more Intel sessions and a booth to follow. The appearance is confirmed. A named customer, a ship date, or a measured performance number for the Crescent Island GPU is not.

Sep 7, 2026

Dennis Singleton · Markets Writer

Dennis Singleton has spent years following the markets, but what keeps his attention is how AI is built. He writes about the companies behind the technology, from semiconductor designers and advanced packaging to photonics, memory, networking, and the hardware powering modern AI. His approach starts with filings, earnings, and industry research, then translates the important details into clear, straightforward analysis without unnecessary hype.