OpenAI Releases GPT-5.6 Following U.S. Safety Review

OpenAI is launching its GPT-5.6 artificial intelligence model family on Thursday, July 9, 2026 ^[1][2]. The new series includes three specific models named Sol, Terra, and Luna ^[1][3]. This public release follows the White House lifting a previous request for restrictions after a safety evaluation ^[3].

The U.S. Department of Commerce conducted voluntary national security testing on the new systems ^[1]. These reviews addressed specific concerns regarding advanced coding and cybersecurity capabilities ^[1]. Officials aimed to verify that the models do not present risks to national infrastructure ^[1][3].

GPT-5.6 represents OpenAI’s most advanced technology to date ^[1]. The launch highlights increasing cooperation between artificial intelligence developers and federal regulators ^[2]. Users may begin accessing the new tools across global markets today ^[2][3].