OpenAI has introduced new ChatGPT subscription limits across its consumer tiers, capping the number of paid accounts that can run on a single hardware device or share the same payment card. The measure blocks users from circumventing standard message caps by rotating multiple low-tier subscriptions on one machine.
Key points
- OpenAI now restricts how many individual paid accounts can be registered to one payment card or used on a single device.
- Devices that exceed the threshold display a hard lock warning and cannot run additional paid accounts even after a user logs out.
- The caps cover individual paid tiers from Plus and Go up to Pro 500, while free tiers and corporate workspace plans remain unaffected.
- Users needing higher volume are being directed to official credit top-ups and usage reset purchases instead of buying duplicate plans.
How the new hardware and card caps work
Under the updated policy detailed in OpenAI support documentation, the system monitors unique device fingerprints alongside payment methods. If a customer attempts to start or connect too many subscriptions using the same debit or credit card, the transaction fails. Similarly, hardware that exceeds the allowed quota of active paid plans triggers an on-screen notice stating that the device limit has been reached.
Once that limit triggers on a computer, phone, or tablet, the blocked account cannot access paid features on that specific hardware. OpenAI confirmed that simply signing out of an active profile does not release the slot. The device registration remains tied to the hardware profile, preventing quick account switching designed to cycle through fresh prompt allotments.
The rules target a common workaround among heavy users. Because standard flat-rate tiers like Plus offer set allowances that refresh on rolling windows, some developers and power users bought two or three separate subscriptions. By switching between browser profiles or client windows, they maintained continuous access without paying for higher-tier business or enterprise plans.
Affected subscription tiers
OpenAI included its full ladder of personal subscriptions in the new rules: Go, Plus, and the Pro plans tiered at 100, 200, and 500. Unpaid access is exempt from the device counter, so anyone using standard free access will not trigger the lock or face blocking.
Organisational accounts also bypass the restriction. Subscriptions under ChatGPT Business, Enterprise, and Edu agreements do not count toward the device or card thresholds. Those tiers operate through managed administrative dashboards with seat-based billing designed for multi-user workplaces, where shared office workstations and corporate credit cards are standard.
For individual subscribers, however, the change eliminates multi-accounting as an economic alternative. Heavy compute users previously found it cheaper to buy two entry-level subscriptions than to upgrade to expensive high-volume tiers. The device lock closes that pathway by forcing all high-demand activity on a single machine through one primary account.
Steering power users toward official usage resets
Instead of stacking accounts, OpenAI is directing customers who exhaust their model quotas toward official add-ons. Users who hit their message ceiling can pay for usage resets to clear their cooling-off windows immediately or fund their accounts by buying extra credits.
Both options cost more per prompt than buying a secondary flat-rate plan. A supplementary subscription historically gave a user an entire second bucket of messages for a fixed monthly fee. Direct credit purchases and reset fees charge closer to actual compute costs, ensuring high-frequency querying generates incremental revenue rather than running on subsidised flat rates.
The move aligns with broader adjustments across OpenAI’s consumer platform. As earlier reported when examining how ChatGPT usage limits switch to shared token budgets, the company has steadily moved away from unmetered or loosely policed access pools in favour of strict, predictable consumption boundaries.
Why OpenAI is tightening account controls
Running frontier models remains an expensive server operation. When power users daisy-chain entry-level plans, they draw sustained compute during peak hours while paying consumer rates intended for casual, intermittent use. That practice degrades server headroom for standard users and depresses average revenue per unit of compute.
The policy also addresses fraud and payment abuse. Bulk account resellers frequently use single prepaid cards to activate batches of subscriptions before selling credentials on third-party marketplaces. Hard limits per payment card complicate automated subscription farming and limit chargeback exposure.
At the same time, OpenAI has integrated deeper workplace features into its main client. After moves such as when OpenAI connects ChatGPT to Apple iCloud Mail and Calendar, the tool functions increasingly as an everyday professional dashboard. Separating personal users from commercial operations helps drive commercial teams toward formal corporate contracts.
What happens next for subscribers
Subscribers who currently run multiple profiles on a home or office computer will need to consolidate their workflows into a single account. Users who share hardware with family members or colleagues on personal plans may also encounter the limits if several paid accounts touch the same device.
OpenAI has not published the exact numerical ceiling for cards or devices on its support pages, treating the thresholds as anti-abuse parameters. Customers who encounter payment blocks or hardware warnings must wait for administrative review or migrate additional users onto dedicated business workspaces.
