OpenAI will merge ChatGPT and Codex limits: what will change for users

Starting in December, OpenAI will introduce a shared usage limit for ChatGPT and Codex for users on the Plus, Pro, and Business plans. 

The company announced this.

Instead of separate limits for the two services, users will have a single quota that will be consumed depending on the AI model chosen and the complexity of tasks.

How the shared ChatGPT and Codex quota will work

From December, OpenAI will count ChatGPT and Codex usage in tokens — units the AI model breaks information into when processing prompts and generating responses.

The more complex the task and the more powerful the model, the more tokens a user will typically consume. For example, working on a complex project may use a larger portion of the available quota than a simpler prompt.

At the same time, ordinary conversations with ChatGPT in Instant mode will not consume the shared quota. In other words, the new rules will primarily affect the use of other models and the services’ more advanced capabilities.

The company explains the change by merging ChatGPT and Codex into a single environment where users can work on tasks using AI agents.

What happens if a user exhausts the quota

You will be able to check the remaining shared quota and its refresh time in ChatGPT settings under Settings → Usage.

If a user exhausts the available quota, they will have three options: wait for it to refresh, switch to a higher-tier plan, or purchase additional credits to use the features available with them.

OpenAI expects that most users will not need to wait for quota refreshes more often than they do now. The company says it is preparing for the transition so the changes do not disrupt regular work with the services.

Recall that at the end of October, OpenAI announced results in mathematics, computer science, and physics that could relate to solving important scientific problems. However, the company is initially sharing some materials with selected experts and has not yet published the physical results. Researchers criticize this approach and call for open access for independent verification.

Read also: Alphabet lost $500 billion in market value after a report on AI spending and a slowdown in search advertising