OpenAI cuts $200 Pro usage in half, adds a $500 tier

Starting October 30, the $200 ChatGPT Pro plan buys half the ChatGPT Work and Codex usage it buys today at the same price, and OpenAI's case for why that is fine rests on a new model, not a discount. The New Stack reports that the cut arrives alongside a new $500 Pro tier placed above it.

At a glance

• OpenAI added a $500 Pro tier with 25 times the Plus allowance and access to Ultrafast, which generates GPT-6 Astra tokens at up to 8 times standard speed in Codex.
• On the $200 plan, ChatGPT Work and Codex usage falls from 20 to 10 times Plus, and GPT-6 Pro in Chat drops from 200 to 100 messages per week.
• Existing subscribers get 62,500 usage credits, valued by OpenAI at $2,500, that expire December 31, 2026, while Google's and Anthropic's $200 plans still offer 20 times their base allowance.

If you have not been following, a $200 top tier at 20 times the entry allowance has long been the standard among frontier labs. Google's Gemini AI Ultra and Anthropic's top plan both cost $200 a month and carry that 20x, though Google also bundles storage and other benefits. All of these labs sell a $100 plan at 5x too. According to Sottiaux, OpenAI's $200 Pro plan had been closed to new subscribers and reopens together with the new usage rules.

On October 30, the $200 plan drops to 10 times the Plus allowance

The cut hits two meters. Included usage in ChatGPT Work and Codex goes from 20 times the $20 Plus allowance to 10 times. The GPT-6 Pro limit in Chat falls from 200 messages a week to 100. The price stays at $200, and current limits hold through October 29.

After that, existing subscribers get a one-time grant of 62,500 usage credits, which OpenAI values at $2,500. One subscriber posted OpenAI's email to Pro customers on Hacker News. It says the credits show up under Settings > Usage, fall under the Service Credit Terms and expire on December 31, 2026.

There is one concession. OpenAI commits to not bringing back the 5-hour limit on the $200 plan, so you can spend the whole weekly allowance at whatever pace you choose. The New Stack adds that the pace will likely be faster than you want.

The $500 tier offers 25 times Plus and tokens up to 8 times faster

OpenAI describes Pro 500 as its highest usage allowance, at 25 times Plus. It also unlocks Ultrafast, which promises token generation at up to 8 times standard speed for GPT-6 Astra in Codex. Ultrafast for GPT-6.1 Sol is listed as coming soon. Testing Catalog's post on the launch gave the same 25x figure, Ultrafast access and no 5-hour limit on Pro 500.

OpenAI's email says Pro subscriptions now include dots, always-on agents powered by GPT-6 Astra. Each dot has its own cloud computer, learns from feedback over time and can work toward your goals around the clock. Conversations with your dot, and work it does directly, do not draw on your ChatGPT usage allowance.

Sottiaux says the new $200 plan nets out at half the API spend

Tibo Sottiaux, OpenAI's head of product and platform, previewed the change on X before the official announcement, and his post is the clearest statement of what the cut means in dollars:

Tomorrow we are re-opening the Pro $200 subscriptions to new subscribers, but together with it we are also changing how we calculate the usage for it. In effect, if you do the math, it will net out at half the dollar in API spend compared to the old Pro $200 plan.

His argument, repeated in the email, is that cheaper and more efficient models make up the gap. OpenAI says GPT-6.1 Sol brings near-Astra capability to a model you can use every day, and Sottiaux says you will "still get more work done than if you were on the Pro $200 subscription one month ago." Developer replies were negative, with one reading: "Ok guys, back to claude. RIP openAI."

Pro allowances are stated as multiples of the $20 Plus plan

OpenAI quotes both Pro allowances as multiples of Plus. From October 30, Pro 200 buys 10 Plus-sized budgets for ChatGPT Work and Codex, while Pro 500 buys 25. Sottiaux adds a second unit, the dollar value of the same work at API prices, and by that measure the $200 plan's included usage halves.

Think of a carrier selling data plans as "ten times the basic bundle" without ever saying how many gigabytes the basic bundle holds. The transition grant is counted in yet another unit: usage credits with a stated dollar value, rather than a multiple of Plus.

OpenAI never gives Plus as an absolute figure, so every multiple in this story rests on an unstated base. In our view, the $500 tier is a thin step up: 2.5 times the old price buys 25x, only a quarter more than the 20x Pro 200 carried through October 29. As The New Stack suggests, the early preview on X likely shows OpenAI expected this reaction.

Whether 10x at $200 sticks

Several things are still unknown. OpenAI has not given a date for Ultrafast on GPT-6.1 Sol, and it is unclear what the $500 plan's 25x amounts to in practice. The bigger open question is competitive: Google and Anthropic still sell 20x for $200. We do not yet know whether other labs follow OpenAI to 10x, or whether the pushback gets strong enough that OpenAI undoes the change.