OpenAI complied with a US government request by launching its latest GPT-5.6 series through a limited preview to a small group of trusted partners before a wider commercial rollout. The company introduced three models: GPT-5.6 Sol, positioned as its flagship and most capable system; Terra, described as a strong lower-cost option; and Luna, its fastest and most cost-efficient model. OpenAI claimed GPT-5.6 Sol is its strongest model to date, delivering improvements across coding, biology and cybersecurity tasks, and stated it features its most robust safety stack. The limited preview followed engagement with the US administration, with OpenAI sharing its release plans and model capabilities ahead of the launch. “At their request, we are starting with a limited preview for a small group of trusted partners whose participation has been shared with the government, before releasing more broadly,” OpenAI stated. It added participation by the initial partner group had been disclosed to the government but did not identify the entities. OpenAI plans to make the GPT-5.6 models generally available in the coming weeks but framed the restricted release as a short-term move. The approach highlights mounting scrutiny of frontier AI developers after the Trump administration’s recent executive order encouraged companies to allow government assessment of advanced model capabilities before full release. OpenAI stated it is working with the administration on a cyber executive order framework and a repeatable process for future model launches.