OpenAI told developers at DevDay 2026 in San Francisco that ChatGPT has reached 1.2 billion weekly users, and the same keynote introduced a Pro 500 tier priced at $500 a month. Both announcements date to Sept. 29, 2026, a week before the date on many reposts of the news.
The scale figure came as a single line in a presentation of more than 20 announcements. The price was one of the details that changed who gets the top of the plan ladder.
The 1.2 billion weekly users OpenAI reported
OpenAI’s own DevDay 2026 recap lists the user base at 1.2 billion weekly users alongside more than 20 announcements. The figure is OpenAI’s own and describes weekly users across ChatGPT, not paying subscribers. SQ Magazine adds that OpenAI also reported 2.5 million businesses on the platform, and that Sam Altman opened the keynote by emphasizing scale. The number has no outside audit attached to it. It stands as a company statement made on stage and repeated in OpenAI’s written recap, and the 2.5 million business figure carries the same status.
Pro 500 and the reshuffled Pro ladder
The new tier is called Pro 500. OpenAI describes it as offering 25 times the ChatGPT Plus allowance with the new Ultrafast speed tier included, and BetaNews reports its price as $500 a month, a figure the Baltimore Chronicle repeats in its tier list of Pro 100 at $100, Pro 200 at $200 and Pro 500 at $500.
The $200 tier did not stay as it was. SQ Magazine reports that Pro 200 reopened with reduced allowances: Codex and Work usage cut from 20 times to 10 times the Plus level, and GPT-6 Pro chat reduced from 200 to 100 weekly messages, with grandfathering for existing subscribers running through Oct. 29, 2026. OpenAI’s help page for Pro had stated as recently as Sept. 10, 2026 that new sign-ups to the $200 plan were halted. The ChatGPT pricing page now describes Pro as offering a choice of three usage tiers.
Ultrafast is the feature that separates the top tier, and OpenAI prices it as a premium mode: Latent Space’s DevDay roundup lists Astra Ultrafast at $60 per million input tokens and $300 per million output tokens in the API, with generation about six times faster there against up to eight times faster in Codex. BetaNews says it reaches up to 300 tokens per second in Codex, up to eight times faster than standard speed. The community announcement thread at OpenAI’s developer forum describes it as up to eight times faster token generation for certain applications.
Dots on GPT-6 Astra
The marquee product launch was Dots, always-on agents powered by GPT-6 Astra that each run on their own cloud computer with a browser. Users assign goals, connect apps, and set which actions can proceed on their own and which need approval, with Latent Space counting more than 4,000 connectable apps; the agents keep working between conversations and learn from feedback over time. Thurrott’s report describes them as frontier always-on agents that can be reached through ChatGPT, Slack and Teams.
Availability is narrower than the keynote suggested. Dots are rolling out to Pro and Business Premium subscribers in eligible markets, with the Pro release excluding the European Economic Area, Switzerland and the U.K., according to BetaNews. Enterprise, Education and Healthcare workspaces can enable a beta, and the first dot is included at no extra cost. Saachi Jain, OpenAI’s head of safety systems, said a planned variant “didn’t quite meet the bar in terms of staying within scope and authorization,” a limit on what shipped that bears directly on how much autonomy a paid agent is given.
Around Dots sits a set of workplace additions from the same recap: ChatGPT Space for shared team workspaces, Pages for human-and-agent document work, @ChatGPT mentions in Slack and Microsoft Teams, an Agents API that now includes computer use, and Sign in with ChatGPT across 16 partner platforms. Slack and Teams access arrives for Business and Enterprise accounts, according to the Baltimore Chronicle, which places the collaboration features in the paid workplace plans.
GPT-6.1 Sol and the cheaper end
Alongside the premium tier, OpenAI released GPT-6.1 Sol, a model it positions as near-Astra performance on coding and professional work at one-fifth of Astra’s token prices. SQ Magazine lists the API rate at $2 per million input tokens and $10 per million output tokens, and OpenAI’s recap says it is available across the API and the Plus, Pro, Business, Enterprise and Edu plans.
The package of tiers now runs from Plus through Pro 100, Pro 200 and Pro 500, with a model priced for cost-sensitive use at the other end. What OpenAI has not published alongside the 1.2 billion figure is how many of those weekly users pay for any plan, which leaves the commercial weight of a $500 tier unmeasured.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- Comcast is shutting down its email service, and Xfinity customers must act to keep it
- Invisible volcanic gas at a California ski resort has killed three ski patrollers
- Verizon just moved to cut landline phone service in 9 states within weeks
- Tropical Storm Rachel is dumping up to 12 inches on four Mexican states on its way to major hurricane strength
