Sol, Earth and Moon: OpenAI launches new GPT-5.6 models

OpenAI has announced that its new cutting-edge artificial intelligence (AI) model, GPT-5.6will be made available to the public on Thursday, after an initial launch limited to some US partners.

The GPT-5.6 range, created by the ChatGPT manufacturing company and including the models Sol, Earth and Moonwill launch publicly this Thursday, OpenAI said in a statement posted on X. “We are expanding early access globally,” the company added.

Safety concerns and model variants

OpenAI’s new offering, GPT-5.6, along with other advanced AI models, notably Anthropic’s Mythos series, has raised concerns due to its supposedly unprecedented ability to identify software vulnerabilitiesor flaws in the code that hackers can exploit.

The GPT-5.6 series includes three new models: Sol (the company’s new flagship product), Terra (a mid-range model designed for everyday use), and Luna (a fast and affordable option).

The green light from the US administration and the market scenario

The US website Axios reported, citing a source close to the matter, that the Trump administration has given OpenAI the green light for a large-scale launch of GPT-5.6, following tests and meetings between the company and government officials.

This announcement follows one similar development by AnthropicOpenAI’s rival, which last week announced it was restoring global access to its most powerful AI models, Fable 5 and Mythos 5, after the US government lifted restrictions on the regions where they could be sold.