OpenAI launches GPT-5.6 model family after US national security review

OpenAI has publicly launched its new GPT-5.6 model family, consisting of three variants—flagship Sol, intermediate Terra, and budget-friendly Luna—following a delayed rollout attributed to U.S. national-security scrutiny and a voluntary government safety review. The release marks a significant step in the AI industry's maturation under emerging federal regulations, as the Trump administration previously sought to restrict the model due to concerns over potential misuse in cybersecurity and other high-risk domains. Developers and enterprise users can now access the models globally through ChatGPT, Codex, and the OpenAI API, with pricing starting at $1 per million tokens for Luna and reaching $30 for Sol output, while broader availability is expected to complete over the next 24 hours. This launch impacts a wide range of fields, promising enhanced capabilities in coding, scientific research, and enterprise work, while setting a new benchmark as Sol reportedly outperforms competitors like Claude Mythos 5 using significantly fewer tokens.