OpenAI has announced that it will publicly release its latest generation of artificial intelligence models on Thursday, marking a significant expansion of access after previously limiting the technology to a select group of U.S.-based partners.
The company said the GPT-5.6 series will include three models designed for different types of users and workloads. The lineup features Sol, OpenAI’s flagship model, Terra, aimed at general-purpose tasks, and Luna, a faster and lower-cost option intended for lightweight applications.
In a post on X, OpenAI confirmed that the new models would become publicly available on Thursday while also expanding preview access to users around the world. The company did not provide additional details about availability, pricing or supported platforms.
The announcement comes after OpenAI disclosed in late June that it had granted preview access to GPT-5.6 to a limited group of U.S.-only partners at the request of Washington. The restricted rollout reflected growing government interest in evaluating advanced AI systems before they are released on a wider scale.
According to a report by U.S. news outlet Axios, the Trump administration has now approved a broader launch of GPT-5.6 following a series of tests and discussions between OpenAI and government officials. The report cited a source familiar with the matter.
OpenAI has not publicly commented on the report, and the White House and the U.S. Department of Commerce had not issued statements at the time of publication.
The release comes as increasingly powerful AI systems face heightened scrutiny over their capabilities and potential risks. Researchers and cybersecurity experts have expressed concern that the newest generation of models can identify software vulnerabilities with greater accuracy than previous versions. While these abilities may help developers strengthen security, they could also be misused by cybercriminals seeking weaknesses in computer systems.
The debate has intensified as leading AI companies continue to introduce more capable models while governments explore ways to balance innovation with security.
OpenAI’s latest announcement follows a similar move by rival Anthropic. Last week, the company said it would restore global access to its most advanced AI models, Fable 5 and Mythos 5, after U.S. authorities lifted restrictions on where those systems could be deployed.
The developments highlight the rapid pace of competition among major AI developers, with companies racing to deliver more advanced models for businesses, researchers and consumers while navigating increasing regulatory oversight.
Industry observers say the public launch of GPT-5.6 is expected to attract strong interest from software developers, enterprises and content creators eager to test its performance across coding, research, writing and other professional tasks.
As governments continue examining the impact of advanced artificial intelligence, the rollout of GPT-5.6 is likely to become another closely watched milestone in the global AI race.
