OpenAI introduces the new generation GPT-5.6 model family: Meet Sol, Terra, and Luna

OpenAI, a leader in the field of artificial intelligence, has officially announced its latest and most powerful development — the GPT-5.6 model family. This new generation of models is significantly superior to previous versions not only in computing power but also in energy efficiency and cost-effectiveness. With this step, the company aims to take market competition to a new level. This is reported by Techcrunch.com reports .

The GPT-5.6 series is presented in three variants, each specialized for specific tasks. Sol is the most powerful and primary workhorse model. Terra offers mid-level capabilities, while Luna is designed for budget-friendly and fast operations. In an interview with CNBC, OpenAI CEO Sam Altman emphasized that the Sol model is 54 percent more efficient than previous versions when working with programming code.

Cybersecurity and corporate solutions

The most notable aspect of the new model is its capabilities in the field of cybersecurity. According to OpenAI representatives, GPT-5.6 is the most powerful security model in the company's history, delivering high results while consuming fewer resources. This technology assists security professionals in complex processes such as threat modeling, code analysis, and vulnerability detection (blue teaming).

The company also introduced a new tool called ChatGPT Work. This platform is specifically designed for corporate teams and works on desktop, web, and mobile devices. With ChatGPT Work, it is possible to automate daily office tasks such as drafting documents, working with complex spreadsheets, and creating presentations.

Competition and performance metrics

The release of GPT-5.6 comes at a time when competition in the AI market is peaking. With this innovation, OpenAI aims to deal a serious blow to its main competitor, Anthropic. According to the Artificial Analysis Coding Agent Index, the Sol model has outperformed the recently introduced Fable model from Anthropic across all metrics.

Specifically, the Sol model scored 80 points in programming tests, showing a result 2.8 points higher than Fable 5. Most importantly, the OpenAI development achieves this result in half the time and at one-third the cost. This allows for significant cost reductions for corporate clients.

It is worth noting that the release of this model also caused certain political debates in the USA. Earlier, the Trump administration attempted to restrict its public release, fearing the risk of the technology being used for malicious purposes. However, by proving that its security measures are at a high level, OpenAI managed to bring the new generation of models to the global market.