On July 9 (local time), OpenAI made a significant announcement: the complete rollout of the GPT-5.6 series models. This series comprises the flagship GPT-5.6 Sol, the well-balanced Terra, and the cost-effective Luna. Notably, Sol features an Ultra mode, which, by default, can orchestrate four AI agents to tackle complex tasks simultaneously. This capability greatly boosts efficiency in various fields, including code development, scientific research, cybersecurity, and knowledge-based work. OpenAI highlighted that the GPT-5.6 series excels in multiple benchmark tests. It substantially cuts down on inference costs and response times, all while maintaining or even enhancing performance. Moreover, the models in this series are outfitted with the most comprehensive safety protection system available to date. They also officially support programmatic tool invocation, thereby bolstering their automatic execution capabilities for handling complex tasks.