Who is already allowed to use OpenAI's o3 and o3 mini AI models? A phased rollout is planned – including a free version

Konrad Wolfenstein

1 year ago

Who is already allowed to use OpenAI's o3 and o3 mini AI models? A phased rollout is planned – including a free version – Image: Xpert.Digital

AI update: Availability and access to o3 and o3 mini

Currently, o3 and o3-mini are not yet available to the general public. However, OpenAI has set up limited access for certain groups:

1. Security researcher

OpenAI has launched a “Public Safety Testing” program, giving selected security experts access to o3-mini to test the model for potential vulnerabilities and security risks. Interested researchers could apply for this early access until January 10, 2025.

2. Internal Tests

OpenAI is currently conducting its own internal tests to ensure the safety and reliability of the models.

Public Launch Schedule

The planned timeline for the public launch is as follows:

o3-mini: Expected release at the end of January 2025. Sam Altman, CEO of OpenAI, confirmed on January 17, 2025, that o3-mini will be available in about two weeks both as an API and in ChatGPT.
O3: The full O3 model is expected to be released to the general public shortly after the O3 mini. An exact date has not yet been announced, but it will likely be in February or March 2025.

Furthermore, o3-mini is to be integrated into OpenAI's ChatGPT as a free version.

Planned access via API and ChatGPT

Once o3 and o3-mini are available, they will likely be accessible via both the API and ChatGPT. OpenAI plans to launch both access methods simultaneously, in response to user feedback.

Safety and phased implementation

It is important to note that OpenAI places great emphasis on safety and has therefore chosen this phased rollout. Sam Altman, CEO of OpenAI, has even indicated that he would prefer the introduction of a government testing framework to monitor and minimize the risks of new reasoning models before their release.

Related to this:

Summary of what's coming: OpenAI's new AI model "o3 mini" – release in the coming weeks

Discover the strengths of o3 and o3-mini

Both o3 and o3-mini are particularly well suited for complex tasks in various areas:

Science and mathematics

o3 distinguishes itself through outstanding performance in solving complex mathematical and scientific problems.
It achieved an accuracy of 96.7% in the American Invitational Mathematics Examination (AIME), which represents a significant improvement over previous models.

Programming and software development

Both models demonstrate impressive programming capabilities.
o3 achieved an accuracy of up to 71.7% in the SWE-Bench Verified, which represents a significant improvement over previous models.
The models can help with error detection and code optimization.

Data analysis and research

o3 and o3-mini can quickly and accurately analyze large datasets.
They are suitable for analyzing scientific data and can help improve medical diagnoses.

Artificial General Intelligence (AGI)

o3 achieved an accuracy of 87.5% in the ARC-AGI benchmark, indicating progress towards general artificial intelligence.

Efficiency and flexibility

The o3-mini offers three speed levels (low, medium, high) that allow users to adjust performance and efficiency to their specific needs.
It is particularly well suited for applications that require good value for money, such as automated customer services or chatbots.

Education and tutoring

The models can be used as virtual tutors or teaching assistants, especially for complex mathematical and scientific topics.

Enterprise applications

From automated reports to data-driven decisions, companies can significantly increase the efficiency of their operations with these models.

Overall, o3 and o3-mini demonstrate impressive capabilities in areas requiring complex thinking, problem-solving, and data processing, with o3 being optimized for the most demanding tasks and o3-mini for efficient, cost-effective solutions.

The o3 mini offers several advantages for small businesses and start-ups

Cost efficiency

Significantly cheaper than the full version of o3, making it attractive for companies with limited budgets.
Three speed levels (low, medium, high) allow for flexible adjustment to performance and cost requirements.
Even the lowest setting requires fewer computing resources, which reduces operating costs.

Performance

Even the medium speed setting surpasses the performance of the predecessor model o1.
Offers advanced AI capabilities for complex tasks without requiring the full computing power of larger models.

versatility

It is suitable for various application areas such as automated customer service, chatbots, e-commerce recommendations and process automation.
It supports APIs for function calls and structured output, which facilitates integration into existing systems.

Scalability

Flexible and adaptable to growing business requirements without requiring significant financial resources.
It enables startups and small businesses to use high-quality AI solutions and compete with larger companies.

Efficiency improvement

It can be used to automate and optimize various business processes, which increases productivity.
Enables data-driven decisions and improved analyses without significant resource expenditure.