ChatGPT Pro vs Plus Codex limits: what actually differs (November 2026)
As of 2026-11-06, OpenAI's Help Center says Work and Codex usage on Plus and on Business Standard seats may be limited over a five-hour window and a weekly window, and where both apply you need allowance left in both; Pro 100, Pro 200 and Pro 500 currently have no five-hour limit in Work and Codex, though the plan still includes an allowance. The allowance does not correspond to a fixed number of messages, and no multiple between the tiers is worked out here. Since 2026-10-29 Pro 100 and Pro 500 include twice their previous allowance and the Pro 200 reduction is canceled. From December, ChatGPT and Codex share one allowance measured in tokens, and OpenAI has not published an exact date.
Updated 2026-11-07
Switch models on the same endpoint and measure the cost per token yourself; check /models for callable models.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Four official points
Plus has a five-hour window; the three Pro tiers currently do not
Work and Codex usage on Plus and Business Standard may be limited over a five-hour window and a weekly window, and where both apply you need allowance left in both; Pro 100, Pro 200 and Pro 500 still have an included allowance.
Estimated GPT-6 Astra local messages per five hours on Plus
OpenAI's estimates of local messages per five hours for Plus and Business Standard differ by model: GPT-6 Astra 5–45, GPT-6.1 Sol 15–160, GPT-6 Sol 15–150, GPT-6 Luna 350–3,000. OpenAI says these ranges are estimates, not guaranteed message counts; cloud tasks may use more allowance than local messages, and weekly limits may also apply.
From December ChatGPT and Codex share one allowance
OpenAI says that from December the Plus, Pro and Business allowance is shared across ChatGPT and Codex and measured in tokens, and that Instant chat does not draw from it. OpenAI has not published an exact date.
Ultrafast is on the top tier only
Ultrafast uses your included usage first at 8× the Standard rate, and once that allowance is used it draws from your credit balance at 6× the Standard credit rate. OpenAI says Plus, Pro 100, Pro 200 and Business do not include Ultrafast at launch.
Where Plus and Pro Codex limits actually differ
As of 2026-11-06, going by OpenAI's Help Center, the clearest difference between Plus and Pro is the five-hour window described above: Plus and Business Standard may have a five-hour window and a weekly window, and where both apply you need allowance left in both. The allowance does not correspond to a fixed number of messages or tasks: OpenAI gives local-message estimate ranges per five hours by model for Plus and Business Standard and states that the Pro tiers still include an allowance; no multiple is worked out here. Which tier fits depends on whether you are hitting the five-hour window, the weekly window or the included allowance itself.
2026-10-29: Pro allowance doubled, Pro 200 reduction canceled
OpenAI's Help Center update of 2026-10-29 says Pro 100 and Pro 500 now include twice their previous Work and Codex allowance, and the Pro 200 reduction announced in the 2026-09-29 email is canceled, with Pro 200 subscribers on the lower allowance doubled to match the higher one. The full account of that change is on the Pro limit change page under related reading.
Timeline
An OpenAI email announces that Pro 200 will move to a lower Work and Codex allowance; it was later canceled.
OpenAI Help Center update: Pro 100 and Pro 500 allowances doubled, the Pro 200 reduction canceled, and Pro 200 on the lower allowance doubled to match the higher one.
Checked: the windows and estimate ranges on this page were checked against OpenAI's Help Center on this date.
What OpenAI states, and what is left out here
Confirmed (OpenAI Help Center wording)
① The allowance does not correspond to a fixed number of messages or tasks. ② Work and Codex usage is included with Plus and Pro plans and with Business Standard and Premium seats. ③ Depending on the plan, a five-hour window and a weekly window may apply, and where both apply you need allowance left in both. ④ Pro 100, Pro 200 and Pro 500 plans still include an allowance (five-hour window: see above). ⑤ After hitting a limit, depending on your account you may wait for the reset, use a saved reset or continue with eligible credits, and switching models does not restore allowance in a shared usage pool. ⑥ Eligible personal Plus and Pro accounts can buy an instant reset, depending on account and billing country. ⑦ API usage is billed separately at API rates.
What is not written here
No multiple between the Plus and Pro tiers is worked out, and no multiple circulating online is quoted; for the December shared allowance, no pool size per tier or start date is given, only OpenAI's own statements.
Choosing by the limit you hit
You hit the five-hour window
The stated difference is that Plus has a five-hour window and Pro 100, Pro 200 and Pro 500 currently have none in Work and Codex. Pro still has an included allowance. Ultrafast is available only on Pro 500.
You hit the allowance itself
Whatever the tier, the ways out OpenAI lists depend on your account: wait for the reset, use a saved reset or continue with eligible credits; eligible personal Plus and Pro accounts can also buy an instant reset. Once the allowance is shared in December, the listed options are to wait for the reset, upgrade your plan or buy credits. The step-by-step version is on the allowance-used-up page under related reading.
How to choose: find which limit you hit first
① In your own ChatGPT usage page, see whether you hit the five-hour window, the weekly window or the included allowance. ② If only the five-hour window stops you: that is exactly where Plus and Pro differ (see above), though Pro still has an included allowance, and Ultrafast is only on Pro 500. ③ Plan for December: once ChatGPT and Codex share one allowance, your ChatGPT use other than Instant chat draws from the same allowance as Codex, so size your tier for both together. ④ To move part of your Codex work to token-billed API usage, OpenAI states that API usage is billed separately at API rates, and a QCode API key is one such path, described in the next section.
Running Codex with a QCode API key
As of 2026-11-06, QCode's /models lists gpt-6.1-sol, gpt-6.1-terra, gpt-6-sol and gpt-6-astra, billed per token with per-model rates on /models. It runs on an API key and does not raise your ChatGPT subscription allowance. Three steps: ① register and top up (Alipay, WeChat Pay, cards, crypto and more); ② buy a Personal Plan or a Pay-as-you-go Pack and get your API key, which starts with cr_; ③ in the model_providers table of ~/.codex/config.toml write base_url = "https://api.qcode.cc/openai" and wire_api = "responses", set model to a listed model, and see the Codex page on docs.qcode.cc for the full setup. Not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days; if you already have a Personal Plan, add a booster pack when you need more quota.
Frequently asked questions
How much bigger is the Codex allowance on ChatGPT Pro than on Plus?
OpenAI states that the allowance does not correspond to a fixed number of messages, and no multiple is worked out here. The confirmed difference is the five-hour window (see above), and every tier has an included allowance. Your own usage page shows what is left on your tier.
How many Codex messages does Plus give per five hours?
OpenAI gives estimates, not guarantees: local messages per five hours are GPT-6 Astra 5–45, GPT-6.1 Sol 15–160, GPT-6 Sol 15–150 and GPT-6 Luna 350–3,000, the same for Plus and Business Standard. Cloud tasks may use more allowance than local messages, and weekly limits may also apply.
Pro 100 or Pro 200: which should I pick for Codex?
On what OpenAI states, the two are the same: as of 2026-11-06, Pro 100 and Pro 200 currently have no five-hour limit in Work and Codex, both still include an allowance, and neither includes Ultrafast at launch (it is on Pro 500 only). No multiple between their allowances is worked out here, so the practical test is whether the included allowance on your tier runs out before it resets. OpenAI says "currently", so once the allowance becomes shared in December, rely on the official pages.
I hit the five-hour limit on Plus but still have weekly allowance. What now?
Weekly allowance alone is not enough: OpenAI says that where both windows apply you need allowance left in both to continue. Depending on your account you can wait for the reset, use a saved reset or continue with eligible credits; eligible personal Plus accounts can also buy an instant reset, depending on account and billing country; switching models does not restore allowance in a shared usage pool. If the five-hour window keeps stopping you, see above for how the three Pro tiers differ on it. More ways out are on the allowance-used-up page under related reading.
What happens to Plus and Pro Codex usage once the allowance is shared in December?
OpenAI says that from December the Plus, Pro and Business allowance is shared across ChatGPT and Codex and measured in tokens, Instant chat does not draw from it, and no exact date has been published. You will be able to see the combined remaining usage and reset times in Settings → Usage; if you reach the limit, you can wait for the reset, upgrade your plan or buy credits. OpenAI expects most users won't need to wait for a reset more often than they do now.
If I run Codex with an API key, how is it billed?
Separately from your ChatGPT plan: OpenAI says API usage is billed separately at API rates, and an API key does not raise your ChatGPT allowance. With a QCode key it is billed per token, with per-model rates on /models; set it up as on the Codex page of docs.qcode.cc: base_url = https://api.qcode.cc/openai and wire_api = responses in config.toml, with a GPT model that /models lists.
Sources
OpenAI Help Center, three articles: Understanding ChatGPT and Codex usage limits, Managing usage with GPT-6 Astra in Work and Codex, and the ChatGPT Business release notes (entry of 2026-10-29), read on 2026-11-06. QCode: qcode.cc/models and the Codex page on docs.qcode.cc, retrieved on 2026-11-06.
Hit a limit? Find out which one first
One QCode key runs Codex on gpt-6.1-sol, gpt-6-sol and gpt-6-astra, billed per token; per-model rates are listed on /models.
Related reading
Pay-as-you-go Packs: no subscription, 7 days
One-time quota packs valid for 7 days, no subscription; per-model rates are on /models.
The Codex weekly allowance vanished in hours
Where the allowance goes, how reasoning levels differ, and the ways out OpenAI lists.
ChatGPT Pro usage limits in Codex: the Pro 200 cut was canceled (October 2026)
The 10-29 change: Pro 100 and Pro 500 doubled, the Pro 200 reduction canceled, and the Pro 500 Ultrafast rates.
Crypto payment guide
Nine coins, 56 networks, zero platform fee for AI coding API credits.
Statements on this page were checked on 2026-11-06 against OpenAI's Help Center; the official pages prevail, and official changes are not announced here. Your real allowance is whatever the usage page of your own account shows; QCode model availability follows /models.