The restriction, introduced on September 10, applies to new sign-ups for the Pro $200 plan, also called Pro 20X, and to users seeking to move to it from Free, Go, Plus or the lower-priced Pro $100 tier. Existing Pro $200 subscribers can continue using their accounts and their subscriptions will renew normally, OpenAI said in updated support documentation.
The company said the move was intended to protect service quality and maintain broad access to Astra while pressure on its infrastructure remains elevated. Thibault Sottiaux, an OpenAI product leader responsible for core products including ChatGPT and Codex, said the highest-priced Pro tier placed the greatest strain on the company’s systems.
“To make sure our current users have an incredible experience and continued access to Astra, we are going to pause subscriptions to our $200 Pro plan,” Sottiaux said in a post on X. He said OpenAI had chosen the “smallest step” that would allow it to preserve the broadest possible access, while other plans and the company’s application programming interface remained available.
OpenAI has not announced when new Pro $200 subscriptions will resume. The pause does not affect new or existing subscriptions to the $100 Pro tier, which remains available with lower usage allowances than the $200 version. Go and Plus subscriptions also remain open to new customers.
Users who cancel Pro $200 or schedule a downgrade keep access until the end of their current billing cycle. They can reverse that decision before the cycle expires, but once the subscription ends they cannot buy the $200 tier again until OpenAI lifts the pause. Customers who downgrade to Pro $100 face the same restriction on moving back to the higher tier while the suspension is in force.
OpenAI also warned that a failed renewal payment could result in the loss of Pro $200 access if the account is ultimately cancelled. Subscribers billed through ChatGPT may receive instructions to resolve a failed payment, while customers using Apple or Google must follow the relevant app store recovery process.
The capacity measure comes as Astra, OpenAI’s newest flagship model, expands across the company’s products after a staggered launch earlier this month. The model has been positioned for demanding professional tasks spanning coding, research, financial analysis and other complex workflows, increasing the importance of reliable access for subscribers using ChatGPT as a work tool.
The launch initially prioritised selected enterprise customers before a broader rollout to paid ChatGPT users and cloud channels. Chief executive Sam Altman acknowledged problems with the deployment after some Plus and Pro subscribers could not immediately access Astra. The episode highlighted the difficulty of matching a high-profile model release with available serving capacity, particularly when usage rises across consumer subscriptions, developer services and business deployments at the same time.
OpenAI this week also introduced ChatGPT for Financial Services, a specialised product built with Morgan Stanley and Evercore as design partners. The service uses GPT-6 Astra and integrates data from providers including LSEG, PitchBook and Daloopa, with tools intended for investment research, financial modelling and preparation of client materials. OpenAI has said the product carries enterprise controls including encryption, role-based access and audit-log capabilities.
Astra’s release has drawn attention not only for its capabilities but also for the computing resources needed to serve advanced models at scale. Strong demand can translate directly into higher inference loads, particularly on subscription tiers that offer heavier usage and access to premium models and tools.
The Pro $200 tier remains OpenAI’s highest-usage consumer subscription. Its benefits include access to Pro models, Codex, deep research, image generation, memory and file uploads, although specific models may have separate usage limits. OpenAI says those allowances can differ between the $100 and $200 Pro tiers.
The company’s support guidance says reaching a model-specific allowance may temporarily make that model unavailable until the quota resets, without affecting the subscription itself. Support staff cannot manually reset those limits, and users are directed to another available model or to wait for the displayed reset time.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
