Originally published on Kuponaut. No affiliate links: Kuponaut has no partner link or code with OpenAI and earns nothing from this article. The help article, the Codex pricing page and the ChatGPT pricing page were read on 3 November 2026; prices are in USD before tax and can change.
OpenAI's help centre now says that, starting in December, the usage allowance on ChatGPT Plus, Pro and Business will be shared across ChatGPT and Codex. If you use Codex for coding and ChatGPT for everyday work on the same plan, both will draw from one token-based pool. This guide separates what OpenAI has confirmed from what it has not published yet, shows what the plans cost today, and lists five practical ways to make the shared allowance last, using OpenAI's own pages read on 3 November 2026.
What OpenAI has confirmed
The help article "Understanding ChatGPT and Codex usage limits" says that from December the Plus, Pro and Business allowance will be shared across ChatGPT and Codex, as OpenAI brings the agentic features of ChatGPT Work into the main ChatGPT experience.
Usage will be measured in tokens, and how much you use depends on the model and the task: OpenAI says more advanced models and complex tasks generally use more of the allowance. Instant chat will not draw from it. You will see your combined remaining usage and reset times in Settings → Usage, and when you reach the limit you can wait for the reset, upgrade, or buy credits for eligible features.
OpenAI also writes that it expects most users won't need to wait for a reset more often than they do now. That is a forecast, not a guarantee.
What is not published yet
As of 3 November 2026 the help article does not give the size of each plan's allowance, the exact start day in December, the reset windows under the new system, or a credit price. It names only Plus, Pro and Business; Free, Go, Enterprise and Edu are not mentioned.
The Codex pricing page gives a hint of how this already works, because ChatGPT Work and Codex already share usage today. For Plus and Standard Business it estimates local Codex messages per five-hour period by model: 5 to 45 with GPT-6 Astra, 15 to 160 with GPT-6.1 Sol, 35 to 285 with GPT-6.1 Terra and 350 to 3,000 with GPT-6 Luna. Pro plans currently have no five-hour limit, and weekly limits may also apply. OpenAI stresses that these are estimates, not fixed limits, and they describe the current system, not the December one.
What the plans cost today (3 November 2026)
OpenAI's own pages do not show the same numbers in every region. The Codex pricing page lists Go at $8 a month and Pro from $100 (Pro comes at $100, $200 or $500 a month). The ChatGPT pricing page we opened from Turkey showed Go at $6 and Pro from $120. Check the price your own checkout shows before you upgrade. We keep dated prices for other AI tools in our AI tool price index.
| Plan | Price (read 3 Nov 2026) | Codex access | In the shared pool from December? |
|---|---|---|---|
| Free | $0 | Limited (GPT-6 Luna in the desktop app, subject to rollout) | Not mentioned |
| Go | $8 on the Codex page; $6 on the pricing page seen from Turkey | Lightweight coding tasks | Not mentioned |
| Plus | $20 a month | Web, CLI, IDE extension and iOS; GPT-6.1 Sol and GPT-6 Luna | Yes |
| Pro | From $100 on the Codex page ($100, $200 or $500); from $120 on the pricing page seen from Turkey | Everything in Plus, no five-hour limit today, Ultrafast on Pro $500 | Yes |
| Business | Per user; see OpenAI's Business pricing | Codex cloud, workspace credits on flexible pricing | Yes |
Five ways to make the shared allowance last
Watch the meter before December. Open Settings → Usage once a week, and in the Codex CLI type /status to see your remaining limits during a session. OpenAI's own advice is to check every week or two so you know your pace.
Use Instant for everyday chat. Instant chat will not draw from the shared allowance, so quick questions, rewrites and summaries that don't need a reasoning model can stay off the meter.
Pick the model to match the task. On the Codex rate card GPT-6 Luna costs 2.5 credits per million input tokens and 12.5 per million output tokens, against 50 and 250 for GPT-6.1 Sol, which makes Luna 20 times cheaper per token (our calculation). The five-hour estimates point the same way: 350 to 3,000 Luna messages against 15 to 160 with GPT-6.1 Sol on Plus. Use the bigger models for hard problems, not for renaming variables.
Keep Codex tasks small and the context lean. OpenAI's tips are to keep prompts precise, give only the relevant files, keep AGENTS.md short and switch off MCP servers you don't need, because every server adds context to each message. Fast mode uses included limits 2.5 times faster and Ultrafast 8 times faster, and image generation uses them about 3 to 5 times faster, so turn those on only when you need them.
Decide in advance what you will do at the limit. You can wait for the reset, buy credits, or move up a plan. If you mostly code, it is also worth comparing what other agents cost per task; we did that for GitHub Copilot Fusion and Claude Code in Copilot Fusion vs Claude Code: which is cheaper per task?.
Who should care most
Heavy Codex users on Plus feel this first: long agent sessions in December will eat into the same pool you use for reasoning chats, deep research and image generation. Light ChatGPT users who rarely open Codex, or who mostly use Instant, should notice little. Pro users currently have no five-hour limit in Codex, but OpenAI has not said whether that stays once the pools merge.
Business admins should check whether their workspace is on flexible pricing, because that is where workspace credits can be bought once members reach the limit.
FAQ
When do ChatGPT and Codex start sharing usage?
OpenAI's help centre says starting in December 2026. No exact day had been published when we read it on 3 November 2026.
Which plans are affected?
The help article names Plus, Pro and Business. It does not mention Free, Go, Enterprise or Edu.
Does normal ChatGPT chat count against the shared limit?
Instant chat will not draw from your usage, according to OpenAI. Other models are not exempt, and more advanced models and complex tasks use more of the allowance.
What happens when I hit the limit?
You can wait for it to reset, upgrade your plan, or buy credits to keep using eligible features. Plus and Pro users can already buy credits for Codex today without changing plans.
Where can I see how much I have left?
In Settings → Usage, which will show your combined remaining usage and reset times. In the Codex CLI you can also type /status during a session.
