The announcement shows OpenAI is revising plans, models and agent features together to pursue monetisation and improved usability at the same time. [Photo: Shutterstock]

[DigitalToday reporter Jinju Hong (홍진주)] OpenAI introduced a new top-tier subscription plan priced at $500 a month (about 670,000 won). At the same time, it lowered key usage limits on its existing $200-a-month Pro plan, further segmenting plans for high-volume users.

On Sept. 29 (local time), IT outlet Engadget reported that OpenAI unveiled the $500-a-month "Pro 500" plan at its annual developer event, "DevDay 2026", held in San Francisco. Pro 500 offers the highest usage limits among OpenAI's personal subscription products and also includes a new premium speed feature called "Ultrafast".

Ultrafast is a premium inference speed tier for tasks where speed matters. According to OpenAI, it supports generation speeds of up to 300 tokens per second in Codex and ChatGPT Work, and provides up to eight times faster token generation in Codex. Ultrafast is also available via the API. GPT-6 Astra-based Ultrafast is currently available, and a GPT-6.1 Sol version is set to be released later.

The existing $200-a-month Pro plan (about 270,000 won) will have reduced usage. Included usage in Codex and Work will drop from about 20 times the Plus plan level to about 10 times. In ChatGPT, the weekly message limit for GPT-6 Pro will also be cut to 100 from 200. Existing subscribers will keep current limits for a period and then receive a one-time credit corresponding to the revised usage.

OpenAI also introduced a personal AI agent called "Dots". Pro users receive one Dot, and conversations with Dots do not count toward standard ChatGPT usage limits. Dots is an "always-on" agent that continuously performs tasks set by users, and OpenAI designed it for use in work automation and as a personal assistant.

It also unveiled a new model. OpenAI introduced "GPT-6.1 Sol" and said it delivers performance close to its top-tier model, GPT-6 Astra, in coding, computer use and professional work. It said standard input and output token prices are about one-fifth of Astra's level.

The new model will be available broadly. Plus, Pro, Business, Enterprise and Education users can use GPT-6.1 Sol in Codex and Work from that day. ChatGPT support will come later.

OpenAI also stressed that it improved the safety of GPT-6.1 Sol. The company said it was improved to follow user intent and safety constraints more reliably and to more transparently reveal limitations that arise during tasks. It also separately released related safety evaluation results.

This plan overhaul is seen as OpenAI further segmenting the top end of its subscription products targeting developers and professionals who use high-performance AI heavily. OpenAI introduced the industry's first $200-a-month Pro plan in late 2024, and competitors such as Anthropic and Perplexity later launched similarly priced premium subscriptions.

With a new $500-a-month price point emerging, competition among high-priced plans in the generative AI subscription market is expected to intensify. As firms not only increase usage but also promote generation speed and AI agent features as differentiators, competition for professional users who use AI as a work tool is expected to begin in earnest.

Keyword

#OpenAI #Pro 500 #Ultrafast #GPT-6.1 Sol #Dots
Copyright © DigitalToday. All rights reserved. Unauthorized reproduction and redistribution are prohibited.