According to OpenAI, on Wednesday July 8 the company announced the full global release of three new AI models—GPT-5.6 Sol, Terra, and Luna—starting July 9, ending a limited preview previously restricted by U.S. government requirements. CEO Sam Altman posted "Happy building" on X, signaling the launch.
OpenAI first introduced the GPT-5.6 series in June under a restricted release to government-vetted institutions. The company stated that while supporting government collaboration on AI capability assessment, it views such review mechanisms as temporary, not permanent policy, as they limit access for developers, enterprises, security teams, and global partners. GPT-5.6 Sol is OpenAI's most powerful model to date, demonstrating enhanced capabilities in programming, biology, and cybersecurity. The expanded preview access aligns with OpenAI's commitment to broad access strategy.