OpenAI has announced significant updates for ChatGPT, including access to the o3-mini model for all users and the introduction of new AI agents. These updates aim to enhance the user experience across various access options, from Plus to Free.
In a recent announcement, OpenAI’s CEO, Sam Altman, shared that all users, including those using ChatGPT for free, will soon have access to the high-performance o3-mini model. Initially announced in December 2024, the rollout of the more affordable and smaller version, o3-mini, was planned for the end of January. This model promises increased performance and better context understanding.
Currently, Plus subscription users can make 100 requests per day to ChatGPT using the o3-mini model. The limitation on model usage is due to the high operational costs, which the company aims to offset with subscription revenue. However, according to Altman, the Pro subscription has not yet achieved this due to high usage.
In addition to model updates, Altman provided insights into the development of AI agents. The newly launched tool, Operator, will soon be available in the Plus subscription. Another AI agent is expected to be available at launch, though the exact timing is unclear.
OpenAI introduced Operator in mid-January 2025. This tool combines the visual capabilities of GPT-4o with advanced decision-making and can interact with Graphical User Interfaces (GUIs). Operator is the first of several Computer-Using-Agents (CUAs) that OpenAI plans to offer. These agents can autonomously complete tasks for users. Developers will also be able to create their own comparable agents via API.
As an AI agent, Operator can perform tasks online, such as reserving tables, filling out forms, or managing shopping. Currently, only Pro users in the USA can access this feature, allowing them to execute tasks in a dedicated browser and monitor them if desired. Custom Instructions can further optimize processes, and future tasks can be scheduled with the new Tasks feature from ChatGPT.
Another AI bot, Claude, has already introduced a Computer User Mode, enabling autonomous actions in the browser. Many major tech companies are working on providing AI agents. In fall 2024, Google was developing a similar feature under the working title Project Jarvis. Other AI agents include Nvidia’s Eureka Agent, Salesforce’s Agentforce, introduced by CEO Marc Benioff, and Microsoft’s Copilot Vision.
These developments indicate a growing trend in AI technology, where companies are racing to enhance AI capabilities and offer more autonomous solutions for users. With these updates, OpenAI continues to lead in the AI space, providing innovative tools and models that cater to a wide range of user needs.
