OpenAI unveils GPT-5.6 amid US AI regulatory drama

OpenAI released GPT-5.6, a three-model AI suite, on Friday, less than 24 hours after reports that the Trump administration would oversee its staggered rollout. The preview includes Sol, Terra, and Luna, with Sol priced at $5 per million input tokens and $30 per million output tokens. The flagship Sol model is positioned for complex tasks, while Terra targets high-volume work and Luna is designed as a fast, affordable option [1]. Sol's pricing is nearly half that of Anthropic's Claude Fable 5, which costs $10 per million input tokens and $50 per million output tokens [1]. Terra is half the cost of Sol, and Luna is less than half the cost of Terra [1]. Anthropic, founded in 2021 by former OpenAI members, had an estimated valuation of $965 billion in May 2026 [10]. OpenAI stated that GPT-5.6 is trained to refuse prohibited cyber assistance, including when users attempt to disguise their intent or jailbreak the model [1]. The company said Sol is better at helping people find and fix vulnerabilities than reliably carrying out end-to-end attacks [1]. OpenAI dedicated approximately 700,000 A100e GPU hours to automated red-teaming and worked with third-party testers who will continue evaluation for the next two weeks [1]. The release comes amid heightened security scrutiny in Washington. The Trump administration will approve customers on a case-by-case basis during the preview period [1]. OpenAI acknowledged that safeguards may occasionally intervene on legitimate work, particularly in dual-use areas where defensive and offensive activity can initially look similar [1]. The company wrote, "We don't believe this kind of government access process should become the long-term default" [1]. OpenAI described the arrangement as a short-term step toward broader availability while working with the administration to develop a cyber Executive Order framework and a repeatable process for future model releases [1]. The model suite is expected to become generally available in the coming weeks [1]. OpenAI also debuted two additional modes for Sol: a "max" mode for deeper reasoning and an "ultra" mode for leveraging sub-agents [1]. The company said Sol has its most robust safety stack to date, with strengthened protections for higher-risk activity, sensitive cyber requests, and repeated misuse [1].