The new subscription includes an Ultrafast mode generating up to 300 tokens per second.
OpenAI introduced a new Pro 500 subscription plan priced at $500 per month during its DevDay event on September 29, 2026. The new top tier offers the company's highest usage allowance and includes access to an Ultrafast mode. In Codex and Work apps, Ultrafast generates up to 300 tokens per second, and the speed tier is also available to API users.
Alongside the new tier, OpenAI reduced allowances for its existing $200 Pro plan. In Codex and Work, usage limits for $200 subscribers drop from 20 times the allowance of Plus users to 10 times. In ChatGPT, GPT-6 Pro message caps fall from 200 to 100 per week. OpenAI stated that existing subscribers will retain current limits for a period and will receive a one-time credit when the lower limits take effect.
All Pro subscribers will also receive access to a personal AI agent called Dot. Conversations with Dot will not count toward ChatGPT usage limits. In addition, OpenAI partnered with 16 companies, including Notion and Cognition, to allow Plus and Pro users to run models directly through a Sign in with ChatGPT feature without configuring an API key.
The company also launched GPT-6.1 Sol, which it claims matches the coding and computer use performance of its GPT-6.0 Astra model at one fifth of the cost. The model is available starting September 29, 2026, in Codex and Work for Plus, Pro, Business, Enterprise, and Education accounts, with access in ChatGPT scheduled for a later date.
Newsletter
Markets in your inbox, weekly
Latin America-focused analysis, investment themes and the week in finance.
Keep reading