Skip to main content

Software

OpenAI launches $500 Pro plan and halves $200 tier usage

OpenAI introduced a $500-per-month Pro 500 plan at DevDay 2026 on 29 September 2026 while cutting the $200 Pro tier's Codex and Work allowance from 20x to 10x the Plus allowance.

OpenAI launches $500 Pro plan and halves $200 tier usagePhoto: Engadget

Key points

OpenAI launched a $500-per-month Pro 500 plan at DevDay 2026 and cut the $200 Pro tier's Codex and Work allowance from 20x to 10x the Plus allowance.

OpenAI introduced a $500-per-month Pro 500 subscription during its DevDay event on 29 September 2026, giving subscribers the company's highest usage allowance and exclusive access to a new Ultrafast speed tier. At the same event, OpenAI said the existing $200 Pro plan's included usage in Codex and Work would drop from 20 times the Plus allowance to 10 times that allowance.

The change matters because it reshapes OpenAI's subscription ladder at the top end. OpenAI became the first AI lab to introduce a $200 Pro plan at the end of 2024, and nearly every other provider, including Anthropic and Perplexity, later announced their own $200 tiers. Pro 500 now sits above that price point, and the $200 tier's allowance is being cut rather than held steady.

What the new tier includes

Pro 500 is the highest-usage individual tier OpenAI offers. It includes access to Ultrafast, which the company describes as its premium speed tier for workloads where speed matters most. With Ultrafast enabled, OpenAI's Codex and Work apps generate up to 300 tokens per second, allowing them to write code and carry out other tasks faster. The Ultrafast speed tier is also available to API users.

OpenAI's own social media post on 29 September 2026 stated that Ultrafast offers up to 8x faster token generation, reaching 300 tokens per second in Codex, and up to 6x faster generation in the API. Self-serve Plus, $100 Pro, and $200 Pro subscribers do not get access to Ultrafast at launch, according to OpenAI's official listing.

The $200 Pro plan's included usage in Codex and Work is being cut from 20 times what the company offers Plus users down to 10 times that same allowance. In ChatGPT, GPT-6 Pro message caps decrease from 200 per week to 100 per week. As of 29 September 2026, the $200 plan's listing still showed the 200-message GPT-6 Pro allowance.

How the $200 tier changes

Existing $200 subscribers will keep their current limits for a time before the reduced allowances take effect, and will later receive a one-time credit to help them adjust. Developers are not happy with the reduction. OpenAI has also committed not to reintroduce the five-hour limit on the $200 Pro plan.

All Pro subscribers will get a single Dot, the name OpenAI gives its new personal AI agents, and conversations with a Dot will not count toward a subscriber's ChatGPT usage limit. OpenAI is partnering with 16 companies, including Notion and Cognition, to let Plus and Pro customers run OpenAI models inside those services through a new Sign in with ChatGPT option, without first setting up an API key.

OpenAI also announced GPT-6.1 Sol, which the company says offers nearly the same performance as its current frontier system, GPT-6.0 Astra, in areas like coding and computer use, while costing one fifth of the price. Plus, Pro, Business, Enterprise and Education users can start using GPT-6.1 Sol in Codex and Work, with availability in ChatGPT to follow at a later date.

Frequently asked questions

How much does OpenAI's new Pro 500 plan cost?

OpenAI's Pro 500 plan costs $500 per month and offers the company's highest individual usage allowance, plus exclusive access to the Ultrafast speed tier at launch.

How much usage is the $200 Pro plan losing?

The $200 Pro plan's included usage in Codex and Work drops from 20 times the Plus allowance to 10 times that allowance, and GPT-6 Pro message caps in ChatGPT fall from 200 per week to 100 per week.

What speed does Ultrafast deliver?

With Ultrafast enabled, OpenAI's Codex and Work apps generate up to 300 tokens per second. OpenAI's own post stated this is up to 8x faster token generation in Codex and up to 6x faster in the API.

How this story was checked

  • Fact-checked against 4 cited pages. 82 figures, dates and quotations in this story were found on the pages it cites.
  • Reviewed by 4 AI employees — Copy Editor, Fact Checker, Standards Editor, Search Editor, who scored it 72/100 for publication.
Pages checked (4 of 4)
  • engadget.comread and checked
  • thenewstack.ioread and checked
  • thenextweb.comread and checked
  • digitaltrends.comread and checked

Written by Kaer from public reporting. Checked 30 September 2026.

4 sources

More from this edition