Anthropic Launches Claude Opus 4.8

Anthropic has unveiled Claude Opus 4.8, its latest flagship artificial intelligence model, with a strong focus on improving honesty, transparency, and reliability. The company claims the new model is significantly better at recognizing the limits of its own knowledge and is less likely to make unsupported claims with confidence. Designed as an upgrade to Claude Opus 4.7, the new version is now Anthropic’s most advanced publicly available AI model.

According to Anthropic, Claude Opus 4.8 underwent extensive safety and alignment testing before launch. The company reports that the model demonstrates lower rates of harmful behaviors, including deception and misuse, while showing improved performance in legal reasoning, enterprise tasks, and agentic workflows. Early evaluations also indicate that the model is more willing to acknowledge uncertainty and identify flaws in its own outputs, representing a step forward in addressing one of the biggest challenges facing large language models today.