ChatGPT Pro Max Leaked: Is OpenAI Preparing a $500 Monthly Tier?
- OpenAI is facing heavy compute strain for its flagship systems as unreleased subscription references point to a potential $500 monthly tier named ChatGPT Pro Max.
- The discovered references outline a monthly subscription priced at $500, with a parallel figure of $600 appearing in the account's video demonstration to account for value-added tax.
- The primary differentiator in the leaked description points directly to agentic computing rather than casual conversation.
OpenAI is facing heavy compute strain for its flagship systems as unreleased subscription references point to a potential $500 monthly tier named ChatGPT Pro Max. The leak surfaced on Sept. 24, 2026, from leak-tracking account @testingcatalog, operated by Tibor Blaho. The figures show up in internal pricing references, though OpenAI has issued no public announcement, launch date, or model identifier for the plan.
Unreleased Subscription References and Pricing Structure
The discovered references outline a monthly subscription priced at $500, with a parallel figure of $600 appearing in the account’s video demonstration to account for value-added tax. Beyond the raw pricing string, the leak lacks concrete launch timelines or confirmed quota allocations. According to the tracking report, the description for the proposed tier differs from the standard Pro plan primarily through access to „Fastest Work and Code“? The discovery arrives just ahead of OpenAI’s developer event scheduled for Sept. 29, 2026. Industry publications including Crypto Briefing, Futubull, and Startup Fortune have picked up the trail, though reporting consistently classifies the details as an unconfirmed leak found within developer tools rather than an official product rollout.”
Targeting Agentic Workloads and Codex Usage
The primary differentiator in the leaked description points directly to agentic computing rather than casual conversation. OpenAI separates chat usage from ChatGPT Work and Codex, with the latter two sharing a single processing pool distinct from standard chat allowances and outside API keys. A plan centered on accelerated work and Codex functions targets long-running agentic jobs. These autonomous tasks represent the highest variable expense for frontier models, where hourly compute costs fluctuate wildly depending on task complexity and execution length. Lower-tier alternatives remain available for developers whose projects do not require dedicated monthly licenses. API options include GPT-6 Sol priced at $2 / $10 per million tokens, alongside GPT-6 Luna priced at $0.10 / $0.50 per million tokens.
Capacity Constraints and the Pause on Pro Subscriptions
The emergence of a potential $500 tier follows strict capacity bottlenecks surrounding the company’s existing flagship models. Around Sept. 11, 2026, OpenAI halted new sign-ups and upgrades for its $200 monthly Pro plan due to overwhelming demand for GPT-6 Astra, which reached general availability on Sept. 5, 2026, and appears internally as GPT-6 Pro. The head of core product and platform stated that the company had never witnessed demand on this scale. The freeze on the $200 tier occurred despite OpenAI expanding its available computing capacity by roughly 9.5 times since 2023. By late September, new users seeking paid access were steered toward a recently introduced $100 Pro tier instead of the paused $200 option. The sequence of events demonstrates a cascading series of pricing tiers adjusted to manage severe infrastructure strain caused by high-demand frontier models.