SAN FRANCISCO, June 27 — OpenAI has begun a limited preview of its new GPT-5.6 family of artificial intelligence models, introducing three variants—GPT-5.6 Sol, GPT-5.6 Terra and GPT-5.6 Luna—while restricting initial access to a small group of trusted partners. The phased rollout comes at the request of the U.S. government as authorities develop a framework for evaluating advanced AI systems with potential cybersecurity implications.
GPT-5.6 Sol is the company’s flagship frontier model, designed for advanced reasoning, coding, cybersecurity and scientific research. It includes new “Max” and “Ultra” operating modes that enable deeper reasoning and more complex agentic workflows. OpenAI said Sol incorporates its most comprehensive safety safeguards to date and has undergone extensive security testing to reduce the risk of misuse.
The company also introduced GPT-5.6 Terra, a mid-tier model optimized for everyday enterprise workloads, and GPT-5.6 Luna, a lower-cost model intended for high-volume applications requiring fast response times. OpenAI said Terra delivers performance comparable to previous flagship models at significantly lower cost, while Luna is designed for scalable deployments.
OpenAI said it previewed the models with U.S. government officials before launch and agreed to limit availability during an initial evaluation period. Under the current arrangement, access will be granted on a case-by-case basis to selected organizations before a broader commercial rollout planned in the coming weeks.
The restricted launch follows growing government scrutiny of frontier AI models capable of identifying software vulnerabilities or performing advanced cybersecurity tasks. OpenAI said the temporary access limits are intended to support ongoing safety evaluations but emphasized that broader availability remains its long-term objective.
The announcement reflects increasing collaboration between AI developers and regulators as governments seek to balance rapid technological innovation with national security and public safety concerns.
Source: Bloomberght/ Prepared by: İlayda Gök