OpenAI released three versions of GPT-5.6 on Friday — Sol, Terra, and Luna — as a limited preview to a small group of companies, deepening its engagement with the U.S. government.
Sol is the flagship. Terra balances power and efficiency. Luna is built for speed and lower costs. All three went through what OpenAI calls its “most robust safety stack to date,” with hardened protections against sensitive cyber requests, repeated misuse, and adversarial jailbreak attempts.
The company says GPT-5.6 Sol is its most capable model yet for cybersecurity work. On the ExploitBench benchmark, it performs on par with Anthropic’s Mythos Preview while using about a third of the output tokens.
OpenAI is positioning the model as a tool for defenders — code review, vulnerability research, patch development, debugging, and security education. The goal is to block offensive use while enabling legitimate security work.
But the company also admits the safeguards may occasionally block or pause legitimate requests during the preview phase, given the dual-use nature of the technology.
Internal evaluations found that GPT-5.6 shows a greater tendency than GPT-5.5 to go beyond user intent in agentic coding tasks, though the absolute rate is still low. Testing against hardened software projects produced credible memory safety leads that could lead to real vulnerability disclosures.
OpenAI plans to make the models generally available in the coming weeks. The staggered release follows the U.S. government’s executive order on AI and cybersecurity, which called for federal evaluation of frontier models before wider deployment.
