ChatGPT adds a $500 Pro plan and halves the $200 plan's allowance

OpenAI now sells ChatGPT Pro in three monthly tiers: Pro 100 at $100, Pro 200 at $200 and a new Pro 500 at $500 [1]. The new tier has the highest included usage of the three and is the only Pro plan with Ultrafast, OpenAI’s premium speed tier [1]. At the same time, the $200 plan keeps its price but buys less: new Pro 200 subscriptions that are not eligible for grandfathering get a lower usage allowance, and existing subscribers move to that lower allowance after 29 October [1]. If you pay for Pro 200 today, your bill does not change, but what you get for it does.

What was announced

OpenAI announced the Pro 500 plan at DevDay on 29 September as one of more than 20 announcements across ChatGPT, Codex and its models [3]. The company describes it as a plan for “those who want even more usage and speed,” with “our highest usage allowance at 25 times the ChatGPT Plus allowance” and access to Ultrafast [3]. It is available now [3].

The help page sets out the three tiers side by side. Pro 100 costs $100 a month and does not include Ultrafast. Pro 200 costs $200 a month and does not include Ultrafast. Pro 500 costs $500 a month and does [1]. All three include Pro models, Codex, deep research, image creation, memory and file uploads, the company says [1]. Pro 200 includes more usage than Pro 100, and Pro 500 offers the most of the three [1]. OpenAI points to its pricing page for the full comparison of features and usage in each plan [1]. Pro is billed monthly; annual billing and paying several months in advance are not available [1].

Ultrafast is the reason the top tier exists. OpenAI calls it “our premium speed tier for workloads where speed matters most” and says it offers up to 8x faster token generation, or 300 tokens per second, in Codex, and up to 6x in the API [3]. GPT-6 Astra Ultrafast is available today in the API and in ChatGPT Work and Codex on Pro 500 and Enterprise plans; a GPT-6.1 Sol version is “coming soon” [3]. Engadget reports the same 300-tokens-per-second figure for Codex and Work [2]. In ChatGPT, Pro 500 users select Ultrafast in the model picker; it draws on the plan’s included usage first and then on the user’s credit balance [1]. At launch, buying credits on Pro 100 or Pro 200 does not unlock Ultrafast [1].

For developers, Ultrafast is not cheap. The Decoder reports that Ultrafast in the API costs six times the standard rate, which for GPT-6 Astra works out to $60 per million input tokens and $300 per million output tokens [5]. That is an API price, separate from the subscription, but it gives a sense of how OpenAI values the speed: Pro 500 bundles some of it into a flat fee.

The Pro 200 change is the part that reaches the most people. OpenAI’s help page says new Pro 200 subscriptions are available again, but that new subscriptions that are not eligible for grandfathering “include a lower usage allowance than previously offered with Pro 200 to reflect our increasingly efficient models” [1]. The monthly price stays at $200 [1]. OpenAI’s page does not quantify the reduction. Engadget does: in Codex and Work, according to Engadget, Pro 200 included usage drops from 20x the Plus allowance to 10x, and in ChatGPT, GPT-6 Pro message caps fall from 200 to 100 per week [2]. Treat those figures as Engadget’s reporting until OpenAI’s pricing page confirms them.

What changed, and for whom

If you already pay for Pro 200, the date to remember is 29 October 2026. According to OpenAI, if your Pro 200 subscription was active at the eligibility cutoff or during the seven days before it, you keep your previous included usage allowance through Oct 29, 2026, as long as the subscription stays active [1]. After that date, the subscription moves to the lower allowance, and the price stays at $200 a month [1]. Only users affected by the change receive an email with details [1]. The help page does not state the eligibility cutoff date itself. Engadget reports that existing subscribers will later receive a one-time credit to help them use the new, reduced allowance [2]; the size of that credit is not public.

Keeping the old allowance does not upgrade anything. OpenAI says a grandfathered subscription “remains Pro 200” and does not add Ultrafast or Pro 500 features [1]. If your subscription lapsed in the seven days before the cutoff, you can still get the previous allowance through 29 October by subscribing again; anyone not eligible gets the updated allowance [1].

If you hit limits, the rules are worth knowing before the change lands. Model allowances vary by tier, and at a model’s limit it may be unavailable until the allowance resets; ChatGPT shows the reset time when it can [1]. Support cannot reset ChatGPT or Codex limits [1]. For eligible features you can buy credits to keep working after the included allowance runs out, and credits cover Codex, ChatGPT Work, ChatGPT for Word, ChatGPT for Excel and ChatGPT for PowerPoint, depending on plan and account [1]. OpenAI notes that credits do not add usage for every model or feature [1].

The same DevDay brought changes that affect what a Pro plan is worth. All Pro subscribers get a single Dot, OpenAI’s new always-on agent, according to Engadget [2]. OpenAI’s own Dots page says the first dot is included in Pro or Business Premium at no extra cost, and that conversations with it do not count toward ChatGPT usage limits, though tasks it starts in Codex or ChatGPT Work do [4]. The plan also includes an allowance for deeper dot work, with extended limits for the first month after launch [4]. OpenAI is also letting Plus and Pro users spend their plan allowance in 16 partner tools, including Cognition’s Devin, Notion, Vercel, T3, OpenClaw and Dactyl, and control how much each can use; eligible usage counts toward plan limits [3]. That last point matters: if you connect partner tools, they draw from the same allowance that Pro 200 is losing.

Who it is for — and not

Pro 500 is for people who already run out of Pro 200 and for whom speed changes the work. If you run long Codex sessions or agents in ChatGPT Work every day and wait on output, the combination of 25x Plus usage and Ultrafast is what the extra $300 a month buys [3]. If Engadget’s 10x figure for the new Pro 200 is right, Pro 500 offers two and a half times the Pro 200 allowance for two and a half times the price, so the price per unit of usage is the same and the speed tier is the extra [2][3]. That comparison rests on two different sources, and OpenAI has not published the Pro 200 multiple itself.

It is not for most people who pay $200 today. Engadget points out that OpenAI introduced the first $200 Pro plan at the end of 2024 and that nearly every other provider, including Anthropic and Perplexity, followed with $200 tiers of their own [2]. A $500 tier could set the next reference price in the same way; whether competitors follow is not known. For a solo operator or small team, the practical question is simpler: how much of your current allowance do you use in a normal month? If you rarely hit a limit, the Pro 200 cut may not change anything you notice, and Pro 100 remains available. If you hit limits every week, the cut will bite after 29 October, and you should decide between paying for credits, moving to Pro 500, or moving heavy work to the API, where you pay per token.

It is also not a reason to rush. Existing Pro 200 subscribers keep their allowance through 29 October [1]. That leaves a month to measure. Pro is billed monthly; annual billing is not available [1]. Cancelling does not end access immediately; you keep the plan until the end of the current billing period and can undo the cancellation before then [1].

This page will be updated if OpenAI’s pricing page publishes the new Pro 200 allowance, the eligibility cutoff date or the size of the one-time credit, and again after 29 October when the change takes effect.

∴ Same $200 price, half the usage after 29 October: check what you actually use before you upgrade or cancel.