ChatGPT Pro 500: Price, Limits and the Pro 200 Cut
OpenAI launched a $500 ChatGPT Pro tier and halved the $200 plan's allowance. We run the arithmetic on the reported numbers and flag where coverage disagrees.
ChatGPT Pro 500 is a new $500 a month plan that OpenAI announced at DevDay on 29 September. It includes the Ultrafast speed tier for GPT-6 Astra, and it arrives together with a cut to the $200 plan: from 30 October, Pro 200's Work and Codex allowance drops from 20 times to 10 times the Plus allowance, according to The Next Web.
What ChatGPT Pro 500 includes
The Next Web reports that Ultrafast on this plan generates up to 300 tokens per second in Codex, up to eight times faster than standard. The Decoder adds that Ultrafast access requires either an Enterprise plan or Pro 500. The Runtime Wire keynote roundup describes the allowance as 25 times Plus usage.
What changes on Pro 200
From 30 October 2026, the Work and Codex allowance falls from 20x to 10x Plus.
GPT-6 Pro messages in chat fall from 200 to 100 per week.
Existing subscribers keep their current limits through 29 October and receive a one-time grant of usage credits worth $2,500, which expires at the end of 2026.
OpenAI said it will not bring back Pro 200's five-hour limit.
Outlets differ on who is hit and when. Runtime Wire says Pro 200 was reopened to new subscribers with a reduced allowance, while The Next Web frames the 30 October date around existing subscribers. Read the notice on your own account before assuming either.
The arithmetic
Using only the reported multiples, here is what each plan costs per unit of Plus allowance:
Plan | Price per month | Work and Codex allowance | Price per 1x of Plus |
|---|---|---|---|
Pro 200 before the cut | $200 | 20x | $10 |
Pro 200 after the cut | $200 | 10x | $20 |
Pro 500 | $500 | 25x | $20 |
After the change, the two Pro tiers cost the same per unit of allowance, so the higher plan no longer carries a bulk discount. The caveat is large: a multiple of Plus is a relative unit, not a published token count, and Pro 500 also buys speed that Pro 200 does not have. Treat the table as a framing device, not a measurement.
Three kinds of user
The cut lands differently depending on where you sit, measured in the same Plus-multiple unit the coverage uses.
Light use, well under 10x. Nothing changes in practice. The plan you have is still more than you spend.
Moderate use, between 10x and 20x. This is the group that feels it. Your options are to pay the higher tier at the same unit price as the new Pro 200, move bulk work to pay-per-token, or trim what you run.
Heavy sustained use above 20x. Pro 500 is the only subscription with headroom, and at that scale the API is worth pricing out as well.
Most people guess their tier from how heavy use feels, not from a measurement, so the first step is to find out which group you are in.
How to decide
Check your real usage before 29 October. If you never came near the old Pro 200 ceiling, a halved allowance may not touch you.
Decide whether speed is your bottleneck. Ultrafast is sold at a premium in the API as well, as we explained in our report on OpenAI Ultrafast.
Compare with pay-per-token for heavy workloads. The model behind the cheaper tier is covered in our GPT-6.1 Sol launch report.
Allowance cuts are becoming a pattern, and we looked at an earlier one in what the Claude Code weekly limit cut means. If you resell AI work or build it into a product, the cost side of the equation is covered in our AI monetization strategies guide, and how to judge limits you cannot see is in how to judge freemium AI agent usage limits.
How did this land?
About the author
Growth & SEO Lead
Manuele covers distribution: SEO, content strategy, and how AI-built products find their first thousand users. He tests everything he recommends.
