OpenAI Launches GPT-5.6 Family with Cybersecurity Focus and Cost Efficiency
OpenAI unveils its newest family of AI models with GPT-5.6, releasing in three distinct variants: Sol as the primary workhorse, Terra as an intermediate option, and Luna as a budget-friendly choice. CEO Sam Altman claims the new models deliver dramatic improvements in efficiency, with Sol using 54% fewer tokens on coding tasks compared to previous versions. The company touts GPT-5.6 as its strongest cybersecurity model to date, supporting defensive capabilities like threat modeling, code review, patching, and blue team simulations.
Alongside the new models, OpenAI introduces ChatGPT Work, a workplace companion tool available across desktop, web, and mobile platforms. The tool targets enterprise teams by helping with daily clerical tasks such as drafting documents, creating spreadsheets, and building presentations. The release positions OpenAI to better serve business customers who increasingly rely on AI for routine productivity work.
The launch intensifies OpenAI's rivalry with Anthropic, which has gained traction as a favored choice among enterprise customers. OpenAI cites the Artificial Analysis Coding Agent Index to claim its models outperform Anthropic's offerings across the board, with Sol scoring 80 on the benchmark — 2.8 points above Anthropic's Fable 5 — while using less than half the output tokens, completing tasks in less than half the time, and costing roughly one-third less.