ChatGPT and Codex Consolidate Usage Limits
agents openai
| Source: HN | Original article
Starting in December, OpenAI will combine usage allowances for ChatGPT and Codex across Plus, Pro, and Business plans.
OpenAI announced that, from December, the token allowances for its ChatGPT and Codex services will be combined into a single pool for Plus, Pro and Business subscribers. The change, outlined in the company’s Help Center, means that usage of the conversational model and the code‑generation model will draw from the same token balance, measured in the units each model processes and generates.
The merger is tied to OpenAI’s rollout of “agentic” Work capabilities inside the main ChatGPT interface. By unifying the limits, the firm aims to simplify budgeting for users who switch between chat and coding tasks, while also ensuring that heavy model use or complex workflows deplete the allowance more quickly. Instant chat remains unlimited, and users will be able to monitor their remaining tokens in the settings menu, wait for periodic resets, upgrade their plan or purchase extra credits if they hit the cap. OpenAI expects the shift to cause only minimal disruption for most customers.
The move matters because many developers and enterprises rely on Codex for code assistance and now must account for those calls against the same quota that powers their conversational queries. A shared limit could alter cost calculations, especially for teams that heavily leverage both modalities, and may prompt adjustments in how OpenAI prices its higher‑tier plans.
Watch for further guidance from OpenAI on the exact reset schedule, any refinements to the unlimited instant‑chat exception, and how the company will handle edge cases where token consumption spikes. User feedback and usage data in the first weeks will likely shape whether the unified model evolves into broader quota reforms across OpenAI’s product suite.
Sources
Back to AIPULSEN