OpenAI unveils GPT-5.6... new models with advanced capabilities in programming and cybersecurity

OpenAI announced the launch of a limited preview of a series of GPT-5.6 models, in a move that represents a new stage in the development of artificial intelligence technologies, with a clear focus on improving programming capabilities, cybersecurity, complex reasoning, and executing long-term tasks. Range.

The announcement comes after reports indicated that the launch was postponed for a short period to conduct additional security reviews, which reflects the growing interest in ensuring the safety of advanced models before making them available to users.

The new series aims to provide higher performance for developers, researchers, and companies, while improving efficiency and lowering costs of use compared to some competing models.

Three models that meet the needs of Different

The GPT-5.6 series consists of three main models designed to meet diverse needs.

The Sol is the most powerful and advanced model, targeting complex tasks that require a high level of thinking and reasoning.

The Terra is designed for business environments and large organizations that require powerful performance with a lower operating cost.

On the other hand, it focuses Luna is applicable to everyday uses, as it provides higher response speed and lower operating cost, making it suitable for commercial applications and services that depend on a large number of requests.

Major improvements in programming and cybersecurity

OpenAI confirmed that GPT-5.6 provides a significant improvement in writing code, Analyze them, discover errors, and suggest more accurate solutions compared to previous versions.

The model also has advanced capabilities in analyzing security vulnerabilities, and helping researchers discover weak points within systems and software, with a focus on defensive uses aimed at enhancing digital security.

The company added that the new models can maintain the context of the conversation and focus during long and complex tasks, which improves the performance of artificial intelligence agents who are able to implement multi-step projects more efficiently.

Versions Additional to the Sol model

Besides the basic version, OpenAI has unveiled two additional versions of the Sol model.

The first version is called Sol Max, and is intended for tasks that require deeper reasoning and more complex analysis.

The Sol Ultra relies on sub-AI agents working together to execute complex projects, allowing tasks to be broken down and analyzed more efficiently. And speed.

The company believes that these versions will be suitable for research institutions, technology companies, and developers working on advanced applications.

Competitive pricing compared to competitors

OpenAI explained that the cost of using the Sol model is 5 per million input codes and $30 per million Output code.

The company confirmed that these prices are about 50% lower than some competing models on the market, while the Terra model comes at a lower cost than Sol, while Luna is the most economical option, making it suitable for large-scale applications that require reducing operating costs.

This step reflects OpenAI’s endeavor to make advanced models more accessible to companies and developers.

Unprecedented security enhancements

OpenAI devoted a large portion of its announcement to talking about the security measures built into GPT-5.6.

It confirmed that the new models were trained to refuse to provide assistance in malicious cyber activities, even if the user tries to hide his intentions or use circumvention methods to obtain information that may be used to carry out cyber attacks.

The company also added advanced mechanisms to monitor attempts at misuse, while improving the model’s ability to distinguish between legitimate security research and offensive activities.

Intensive testing before Launch

OpenAI explained that it conducted extensive testing of the new models before launching them.

These tests included the implementation of hundreds of thousands of processing hours using graphics processing units, in addition to automated attack tests and cooperation with cybersecurity experts and third parties to assess the level of security and discover any potential weaknesses.

The company confirmed that these tests will continue during the limited preview period with the aim of improving the performance of the models before the full launch.

Cooperation with the US government

OpenAI indicated that it coordinated with the US government before launching GPT-5.6, as part of security review procedures aimed at ensuring that models comply with safety standards for advanced artificial intelligence.

At the same time, the company stressed that these reviews should remain temporary, while working to establish a clear regulatory framework that will allow future models to be launched without undue delay. Essential.

The future of GPT-5.6

GPT-5.6 represents a new step in the race to develop artificial intelligence, especially with a focus on improving programming, cybersecurity, and executing complex tasks via artificial intelligence agents.

Technology experts believe that new models may contribute to raising developer productivity, accelerating the discovery of security vulnerabilities, and developing smarter applications, with the continued need to achieve a balance between innovation and ensuring the safe and responsible use of modern technologies.

In an accelerating digital world, the impact of these innovations extends from following Arabic websites and specialized blogs, in addition to what Egypt storesr And Kuwait stores And vitamin stores And foreign sites, relying on A famous website for social services is an essential factor for success and continuity..