OPENAI FIRES BACK AT ANTHROPIC WITH GPT-5.6 — THREE NEW MODELS THAT BEAT CLAUDE ON CODING, COST LESS, AND THE GOVERNMENT TRIED TO STOP IT OpenAI just launched GPT-5.6, a new family of three models called Sol, Terra, and Luna. Sol is the flagship, and OpenAI is not being shy about the fight it is picking. The company went straight at Anthropic, citing the Artificial Analysis Coding Agent Index to claim Sol sets a new state of the art at 80, sitting 2.8 points above Claude Fable 5 while using less than half the output tokens, taking less than half the time, and costing about one-third less. Terra clocks in just above Fable 5 as well. Luna beats Opus 4.8. That is not a subtle message. What makes this release genuinely interesting beyond the benchmark war is the backstory. The Trump administration previously tried to restrict GPT-5.6’s rollout, and OpenAI pushed back. Sam Altman told CNBC that Sol is 54 percent more token efficient for coding tasks. The company also called it the strongest cybersecurity model it has ever built, with native support for threat modeling, code review, blue teaming, and patch generation. In the same breath, OpenAI launched ChatGPT Work, a workplace companion for enterprise teams that can draft documents, build spreadsheets, and handle presentations. Sol comes in at five dollars per million input tokens, thirty for output. The race is not slowing down. Keywords: OpenAI GPT-5.6, Sol model, ChatGPT Work, AI coding benchmark