OpenAI Plans to Pool ChatGPT and Codex Quotas by Tokens Starting December
Frontier AI models · AI assistants · AI coding
OpenAI plans to merge Chat, Work, and Codex usage quotas for ChatGPT Plus, Pro, and Business tiers starting in December, metering consumption by tokens based on model choice and task complexity. Instant mode chats will not deduct from the pooled quota. Exact rates and dates remain unannounced.
OpenAI plans to update its quota metering system for ChatGPT Plus, Pro, and Business plans starting in December. Usage quotas for standard Chat, Work agents, and Codex will be combined and tracked via tokens. While ChatGPT Work and Codex currently share quotas, Chat usage will also be pooled into this shared balance starting in December. Consumption will depend on the chosen model and task complexity, with more advanced models and tasks consuming more tokens, although standard conversations in Instant mode will not deduct from the shared quota. Users will be able to check combined balances and reset times on the Usage page, with options to wait for resets, upgrade plans, or purchase extra credits when exhausted. Specific quota amounts, token deduction rates, and exact December rollout dates have not yet been announced, while Free, Go, Enterprise, and Edu plans are excluded from this update.
Key takeaways
- OpenAI plans to combine Chat, Work, and Codex usage quotas for Plus, Pro, and Business tiers into a token-based system in December.
- Quota deductions will depend on model selection and task complexity; Chat usage will reduce remaining Codex and Work allowances, except for Instant mode.
- Users can check remaining pooled quotas on the Usage page; exact December launch dates and specific token caps have not been released.
Source：iThome Read the original ↗
Prepared, checked and automatically published by AI; not individually reviewed by a human. Translation checked by AI.