Summary (tl;dr)

OpenAI has officially launched its new GPT-5.6 family of large language models—Sol, Terra, and Luna—alongside a new AI agent called ChatGPT Work, which integrates advanced coding capabilities via Codex. The release, which follows a period of government-requested limited access, marks a significant advancement in AI's ability to perform complex, multi-step tasks across various professional domains.

Essential Background

Prior to this release, OpenAI developed highly influential large language models (LLMs) like GPT-3.5 and GPT-4, which powered the popular ChatGPT chatbot, enabling human-like conversations and content generation. The company also introduced Codex, an AI coding assistant designed to help developers write and understand code. The increasing power of these AI models has led to growing scrutiny and discussions with governments regarding their safe and responsible deployment.

The Full Story

Today, July 9, 2026, OpenAI announced the general availability of its highly anticipated GPT-5.6 models: Sol, Terra, and Luna. The flagship model, GPT-5.6 Sol, is hailed for achieving state-of-the-art results in areas such as coding, knowledge work, cybersecurity, and science, boasting improved efficiency and lower costs. Concurrently, OpenAI introduced ChatGPT Work, an innovative AI agent that incorporates Codex to execute multi-step tasks across web, mobile, and desktop applications, leveraging information from various user apps. The rollout of these powerful new models was initially staggered, with a limited preview period since June 26, 2026, a measure taken at the request of the U.S. government to implement and strengthen safeguards against potential misuse.

Why It Matters

The introduction of GPT-5.6 and ChatGPT Work represents a major leap forward in AI capabilities, particularly in the realm of "agentic AI" – systems capable of performing intricate, autonomous, and multi-step tasks. These advancements are expected to significantly boost productivity in coding and general knowledge work, offering more successful outcomes for the same expenditure or comparable results at a reduced total cost due to increased token efficiency. The U.S. government's involvement in the staggered release underscores the increasing regulatory focus and public concern surrounding the safety and societal impact of increasingly powerful AI models. This trend signifies a new era where AI agents are poised to become integral to diverse professional workflows.

Geographic Location

• San Francisco, California, United States (OpenAI headquarters, location of announcements)
• United States (location of government discussions and requests regarding AI model safeguards)