OpenAI announced on Thursday the release of its latest family of AI models, branded as GPT-5.6. The new suite includes three distinct models: Sol, positioned as the company’s primary workhorse; Terra, an intermediate option; and Luna, designed as a budget-friendly choice. These models are intended to enhance user capabilities across various sectors, including enterprise work, software development, and scientific research.
Efficiency and Cybersecurity Focus
CEO Sam Altman stated that the new models offer substantial improvements in efficiency and cost-effectiveness compared to previous iterations. He noted that Sol is reportedly 54% more token-efficient for AI coding tasks. A key highlight of GPT-5.6 is its advanced cybersecurity capabilities, which OpenAI describes as its strongest yet, achieving frontier performance with reduced token usage. The model’s defensive applications include threat modeling, code review and patching, and blue teaming simulations.
The cybersecurity features of GPT-5.6 have drawn attention, particularly given past concerns from the Trump administration regarding potential misuse. The new models support defensive cybersecurity activities by aiding in threat identification, vulnerability assessment, and system security testing.
ChatGPT Work and Competitive Landscape
Alongside the GPT-5.6 models, OpenAI also introduced ChatGPT Work, a new tool developed for enterprise teams. This workplace companion is accessible via desktop, web, and mobile platforms, aiming to assist with routine administrative tasks such as drafting documents, spreadsheets, and presentations.
The release of GPT-5.6 follows recent announcements from competitors like SpaceXAI and Meta. OpenAI appears to be strategically positioning its new offerings, particularly GPT-5.6, to compete directly with Anthropic, a rival company that has gained traction by focusing on enterprise clients. OpenAI asserts that its new models surpass Anthropic’s offerings, citing the Artificial Analysis Coding Agent Index. The company claims Sol, its top coding model, outperforms Anthropic’s Fable 5 by 80.2 points on this metric, while using less than half the output tokens, taking less than half the time, and costing approximately one-third less.
OpenAI further stated that Terra performs slightly better than Fable 5, and Luna exceeds the performance of Opus 4.8, according to their internal analysis using the Coding Agent Index. The GPT-5.6 family is now available through ChatGPT, Codex, and the OpenAI API, with pricing tiers set per million tokens: Sol at $5 for input and $30 for output; Terra at $2.50 for input and $15 for output; and Luna at $1 for input and $6 for output.
Mitchell Landsberg is the senior reporter for News Raise and focuses on Technology. Mitchell regularly writes about social media platforms and how influencers, industry and general people use them to communicate and make money.
