ChatGPT Work and Codex share one usage pool: Plus limits, Pro, credits and what to do

Anyone who pays for ChatGPT and uses Codex now draws on one meter for both. According to OpenAI's Codex pricing page, ChatGPT Work and Codex "share usage": Work inside ChatGPT uses the same pricing, credits and usage limits as Codex. Ordinary ChatGPT chat is metered separately, and API-key usage is billed at API rates. This guide sets out what is shared, what the current numbers are on Plus and Pro, how credits are priced, and how to make an allowance last. Figures below come from OpenAI's page as we read it on 1 November 2026; the page itself says to check your usage dashboard for current limits and reset times.

Key points

• ChatGPT Work and Codex draw on one usage pool; local Codex messages and cloud chats share the plan allowance, per OpenAI's pricing page.
• Plus and Standard Business get estimated local-message ranges per five-hour period, for example 15-160 on GPT-6.1 Sol and 350-3,000 on GPT-6 Luna; OpenAI calls these estimates, not fixed limits, and says weekly limits may also apply.
• Pro plans ($100, $200 or $500 a month) currently have no five-hour limit, according to the same page.
• Codex used with an API key is billed at API prices and does not use subscription limits; Code Review through GitHub is tracked separately.
• Fast mode uses 2.5x included usage and Ultrafast 8x, relative to Standard mode for the same model.

What to do

• Check where you stand: open the usage dashboard in ChatGPT or Codex. In a Codex CLI session, a third-party guide reports that /status shows usage; we did not verify that command against OpenAI's documentation.
• If you are on Plus and hit limits: use a lighter model for routine edits (the page lists GPT-6 Luna with the largest message range), narrow the files and context you give the agent, and avoid Fast or Ultrafast unless you need them.
• If you run automation or CI: OpenAI's page recommends an API key for shared environments, which is billed at API pricing and does not draw on your subscription.
• If you need more headroom: credits can be bought on Plus and Pro to pay for usage after included limits are used up.

What happened

OpenAI's pricing page for Codex states that ChatGPT Work and Codex are included in the Free, Go, Plus, Pro, Business, Edu and Enterprise plans and that they share usage. Local messages and cloud chats draw from the same plan allowance, and the page adds that limits are shared with other agentic features once pricing for them is effective; it says this currently includes ChatGPT for Excel on Plus and Pro.

A caveat on timing: a help-centre article titled "Understanding ChatGPT and Codex usage limits" was circulating on Hacker News on 31 October, but OpenAI's help site refused our requests, so we could not read it. We also could not establish from the pages we could open when the sharing began. A third-party guide published on 27 September already described Work and Codex as sharing an agentic pool, so the arrangement is at least weeks old rather than brand new. What is current is the detailed table on the pricing page.

A third-party guide from aiidelist stresses a distinction that matters in practice: it is "regular Chat versus agentic work", not simply ChatGPT versus Codex. Regular chat stays on separate limits, while longer multi-step tasks in Work and software tasks in Codex share one allowance. That matches OpenAI's wording.

Background

The five-hour window has had a short but turbulent history. Reporting at the time, based on a post by OpenAI's Codex and ChatGPT engineering lead Thibault Sottiaux on X, says the rolling five-hour limit for Plus accounts across ChatGPT Work and Codex was removed in July 2026 and brought back on 25 August. Mezha (citing 9to5Mac) and Progressive Robot both report this. Mezha attributes to Sottiaux the explanation that without the window, some Plus users could use their whole weekly limit in a few intensive sessions and then wait a long time for it to reset.

Progressive Robot quotes the same post on Pro: "We are for the upcoming months keeping the 5h limit not enabled for Pro $100 and Pro $200 subscriptions." OpenAI's pricing page, as we read it now, is consistent with that and goes further, saying Pro plans "currently have no five-hour limit" including the $500 tier. Neither the page nor the reporting gives an end date for that.

Two limits therefore sit on top of each other for Plus: a five-hour window and a weekly cap. The weekly cap, Progressive Robot notes, never went away. OpenAI publishes no fixed weekly number; the page says only that weekly limits may also apply.

The numbers

On the pricing page, OpenAI lists estimated local messages per five-hour period for Plus and Standard Business, with identical ranges for both:

• GPT-6 Astra: 5-45
• GPT-6.1 Sol: 15-160
• GPT-6.1 Terra: 35-285
• GPT-6 Sol: 15-150
• GPT-6 Luna: 350-3,000

The wide ranges are deliberate. The page says model choice, context, reasoning, tool use, retrieval and caching all affect usage, so prompt length alone is not a reliable estimate, and that cloud tasks may use more of the allowance than local messages. Enterprise and Edu plans without flexible pricing have the same per-seat limits as Plus for most features; those with flexible pricing have no fixed rate limits and scale with credits.

Plan prices on the same page: Free $0, Go $8 a month, Plus $20 a month, Business $20 per user per month, Pro from $100 a month in $100, $200 and $500 tiers.

Credits pay for eligible usage after included limits run out. Output tokens dominate the cost, as this comparison of credits per million output tokens shows:

GPT-6 Astra1250

GPT-5.6 Sol500

GPT-6.1 Sol250

GPT-6 Sol250

GPT-6.1 Terra125

GPT-6 Luna12.5

Credits per 1M output tokens, Standard rate (credits)

	credits
GPT-6 Astra	1250
GPT-5.6 Sol	500
GPT-6.1 Sol	250
GPT-6 Sol	250
GPT-6.1 Terra	125
GPT-6 Luna	12.5

Source: OpenAI Codex pricing page, 1 Nov 2026

For input, the same page lists 250 credits per million tokens for GPT-6 Astra, 50 for GPT-6.1 Sol and GPT-6 Sol, 25 for GPT-6.1 Terra and 2.5 for GPT-6 Luna, with cached input much cheaper (2.5 credits for GPT-6.1 Sol). OpenAI says a typical GPT-5.6 Sol task may use 2-15 credits, and that GPT-5.6 Sol's promotional pricing runs at least through 21 November 2026.

Speed modes change the burn rate. Relative to Standard mode for the same model, Fast mode (GPT-6 Astra) uses 2.5x included subscription usage and 2x purchased credits; Ultrafast (GPT-6 Astra and GPT-6.1 Sol) uses 8x included usage and 6x credits. Image generation uses included limits roughly 3-5x faster on average, depending on quality and size.

What it means

For developers who use both products, the practical change is that a long Work task and a coding session compete for the same capacity. Heavy use of one can leave less for the other, as the aiidelist guide puts it; OpenAI's own wording is that they share one usage pool.

Pro subscribers, currently without a five-hour window, are the least constrained. Plus subscribers are the group most likely to feel the ceiling, and the five-hour estimates show how much the model choice matters: the page's own Luna range is about 67 times its Astra range at the top end (3,000 against 45).

Teams that script Codex should note the split between subscription and API-key use. API-key sessions follow API pricing, do not include cloud features such as GitHub code review or Slack, and take model availability from the API models your key can use.

One related deadline: OpenAI says GPT-5.5 retired from Codex with ChatGPT sign-in on 14 October 2026; API-key sessions were not affected.

What we don't know yet

• When the sharing of Work and Codex usage began, and what the help-centre article on 31 October changed, since we could not open it.
• The weekly limit values for any plan; OpenAI publishes none.
• How long Pro will stay without a five-hour window. The only statement we found is "for the upcoming months", from August.
• One Italian site reported a single token-based allowance for ChatGPT and Codex arriving in December. We could not confirm that against OpenAI's pages and do not treat it as fact.

What to watch next

Watch the pricing page for changes to the Pro five-hour exemption, the end of GPT-5.6 Sol's promotional pricing on or after 21 November, and any move of other agentic features such as ChatGPT for Excel into the shared pool beyond Plus and Pro.

Sources

• OpenAI, Codex pricing (primary)
• Mezha: OpenAI has reinstated the five-hour limit for Codex and ChatGPT Work for Plus
• Progressive Robot: ChatGPT usage limit, 5-hour cap for Plus and Pro
• AI IDE List: ChatGPT Work and Codex now share usage
• OpenAI help centre: Understanding ChatGPT and Codex usage limits (could not be opened)