The company said in a statement via “X,” “The (GPT 5.6) group and (Sol), (Terra) and (Luna) models will become available to the public starting Thursday,” adding, “We are beginning early access to these models for users worldwide.”
GPT 5.6 and other advanced artificial intelligence models, including the Mitos series from Anthropic, have raised concerns about what is described as an unprecedented ability to identify software vulnerabilities – that is, weak points that hackers can exploit to penetrate systems or launch cyber attacks.
OpenAI announced in late June that it had given experimental access to GPT 5.6 only to a limited group of partners in the United States, at the request of Washington.
The GPT 5.6 series includes three new models: “Sol”, which is the company’s flagship model, “Terra”, which is intended for daily use, and “Luna”, which is a fast and affordable alternative.
US news website Axios cited a source as saying that the administration of US President Donald Trump gave OpenAI the green light to launch GPT 5.6 on a large scale, after conducting tests and holding meetings between the company and government officials.
Anthropic, the most prominent competitor to OpenAI, announced last week that it would once again provide global access to its most powerful artificial intelligence models, “Fable 5” and “Metos 5”, after the US government lifted restrictions on places to launch them.
