# ChatGPT and Codex will share one usage limit in December

URL: https://techaiwire.com/articles/chatgpt-codex-shared-usage-limit-december-plus-pro-business/

_By Claire Whitmore - Nov 2, 2026_

**By the numbers:**

- 5-45 - GPT-6 Astra local messages per five hours on Plus, OpenAI's estimate
- 350-3,000 - GPT-6 Luna local messages per five hours on Plus, OpenAI's estimate
- $20 - monthly price of ChatGPT Plus, the cheapest plan affected

OpenAI will make ChatGPT and Codex share one usage allowance starting in December, for its Plus, Pro and Business plans. Codex is OpenAI's coding agent, which can work on a developer's own computer or in the cloud. That matters for developers. A long ChatGPT session on a strong model will soon eat into the same budget they need for coding, and the reverse is true too.

OpenAI's help center spelled out the change, and FourWeekMBA and SmartScope both reported it on October 31. "Starting in December, Plus, Pro, and Business plans' usage allowance will be shared across ChatGPT and Codex," the help center reads, [according to FourWeekMBA](https://fourweekmba.com/ai-chatgpt-and-codex-will-share-one-usage-allowance-in-december/). OpenAI has not named a start day.

## What changes in December

From December, one allowance covers both products on the three paid plans. It is counted in tokens, the small chunks of text a model reads and writes. So the cost of each request depends on the model and the task. Stronger models and harder jobs take a bigger share, both sources report.

Chats that use Instant, one of ChatGPT's model options, stay outside the shared meter. "Instant chat will not draw from your usage," the help center says. SmartScope adds that some chat-side caps, such as the message cap on the GPT-6 Pro model, also sit outside the pool.

OpenAI also tried to reassure heavy users. "We expect most users won't need to wait for their usage to reset more often than they do now," its help center says. It has not published the size of the new allowance. It has also not said whether today's time windows will change.

## How the limits work today

Part of ChatGPT already shares the pool. OpenAI's [Codex pricing page](https://learn.chatgpt.com/docs/pricing) says "ChatGPT Work and Codex share usage." Work is one of the agent features inside ChatGPT, and OpenAI bills and caps it the same way as Codex. December widens that pool to ChatGPT itself.

Limits run on a five-hour window, a weekly window, or both, depending on the plan. When both apply, work stops as soon as either one runs out, FourWeekMBA and SmartScope explain. Pro plans have no five-hour cap in Work and Codex today, according to OpenAI.

Some settings drain the allowance faster. OpenAI says the Fast speed tier uses included usage 2.5 times as fast. Ultrafast, which only the $500 Pro tier and some Enterprise and Edu plans get, counts 8 times on GPT-6 Astra. Image generation uses up limits about 3 to 5 times faster on average.

OpenAI's estimates for Plus show how much the model choice matters. They count local messages per five-hour window, meaning Codex work that runs on the developer's own computer. OpenAI says standard Business seats get the same estimates.

| Model | Local messages per five hours on Plus |
|---|---|
| GPT-6 Astra | 5-45 |
| GPT-6 Sol | 15-150 |
| GPT-6.1 Sol | 15-160 |
| GPT-6.1 Terra | 35-285 |
| GPT-6 Luna | 350-3,000 |

"These ranges are estimates, not guaranteed message counts," the help center warns. FourWeekMBA notes the figures may change once the shared pool starts. GPT-6.1 Terra, the mid-priced model OpenAI [released on October 26](/articles/openai-gpt-6-1-terra-api-price-computer-use/), sits in the middle of the table.

## Which plans are affected

The change names three plans. The prices below come from OpenAI's pricing page.

| Plan | Price | Shared allowance from December |
|---|---|---|
| Plus | $20 a month | Yes |
| Pro | $100, $200 or $500 a month | Yes |
| Business | $20 per user a month billed yearly, or $25 billed monthly | Yes |
| Free and Go | $0 and $8 a month | Not named |

Codex can also run with an API key instead of a ChatGPT sign-in. That route is billed at API rates and is separate from any plan's allowance, OpenAI's pricing page says.

Remaining usage and the next reset time will show under Settings, then Usage, [SmartScope reports](https://smartscope.blog/en/blog/chatgpt-shared-usage-limits-codex-starting-december-purchase-2026/). Codex users can already check a usage dashboard, or type /status in the Codex command-line tool during a session. At the limit, SmartScope lists three choices: wait for the reset, move to a higher plan, or buy extra credits.

## What this means for developers

Chat and code will now compete for the same hours. Until now, ordinary ChatGPT chats did not draw on the Codex allowance. From December, a morning of long GPT-6 Astra chats can leave less room for an afternoon of coding.

A few steps help:

- **Use Instant for quick questions.** It is the one chat option the help center keeps off the meter.
- **Check your pace now.** OpenAI's own advice is to "check the dashboard every week or two." Start in November, so you have a baseline before December.
- **Match the model to the task.** FourWeekMBA works out that Plus gets about 67 to 70 times more GPT-6 Luna messages than GPT-6 Astra ones. Save Astra for the hard problems.
- **Move automation to an API key.** Scripts and automated build jobs that run Codex can use API billing, which leaves every plan's allowance alone.
- **Warn your team if you run a Business plan.** Standard Business seats share the Plus estimates, so a heavy chat user can run out of Codex time too.

The open question is size. OpenAI has not said how big the shared allowance will be, or which day it starts. Heavy Codex users should watch the help center in late November for the real limits.

## Sources

- [ChatGPT and Codex Will Share One Usage Allowance in December](https://fourweekmba.com/ai-chatgpt-and-codex-will-share-one-usage-allowance-in-december/) - FourWeekMBA
- [ChatGPT and Codex Usage Allowance Shared From December: Plus, Pro and Business, Token-Based, and What to Do at the Limit](https://smartscope.blog/en/blog/chatgpt-shared-usage-limits-codex-starting-december-purchase-2026/) - SmartScope
- [Pricing](https://learn.chatgpt.com/docs/pricing) - OpenAI