Anthropic released Claude Opus 5 on Friday, describing it as its "safest model yet" and "most aligned model to date." The new model is now the default on Claude Max, Anthropic's highest subscription plan starting at $100 a month, and the strongest available model on Claude Pro, the $20-per-month tier. Anthropic says Opus 5 comes close to Claude Fable 5 in intelligence at half the price and is designed for everyday use. Compared to prior models like Opus 4.8, Sonnet 5, or Fable 5, Opus 5 has the lowest rates of deceptiveness and is least likely to be tricked into misuse. In a test involving a problem the model has never seen, Opus 5 was three times as fast as the next-best model and scored nearly four times higher than OpenAI's GPT-5.6 Sol. Opus 5 performs better or the same as Fable 5 in most Anthropic benchmarks, except health and legal, and beats OpenAI's GPT 5.6 Sol in every benchmark except Agentic coding. Anthropic calls it the "most capable generally available model for scientific research," though Mythos 5 beats it in areas like biology research. Cybersecurity tasks were purposefully left out of Opus 5's training, making it worse at exploiting vulnerabilities than Mythos and previous models. Claude Opus 5 is available as of Friday July 24,