On July 9, 2026, OpenAI began the general rollout of its new-generation LLM lineup "GPT-5.6" (Sol, Terra and Luna), built to strengthen agentic work, across ChatGPT, Codex and the API. The flagship Sol posted a top score of 53.6 points on "Agents' Last Exam (ALE)," a long-horizon professional-work benchmark created by UC Berkeley, and according to the official announcement it became the only model to exceed 50%.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.