Claude Opus 4.8 Launches with Sharper Judgment and Greater Autonomy

Anthropic has officially launched its latest AI model, Claude Opus 4.8, less than two months after the release of Opus 4.7. The new model is described by the company as a "more efficient collaborator," signaling an accelerated iteration pace. Notably, Opus 4.8 keeps the same pricing as its predecessor while enabling a default "high effort" mode, delivering stronger performance on tasks like coding with similar token consumption.

According to Anthropic, Opus 4.8 shows significant improvements in judgment, self-awareness, and autonomous work capability. Early testers report that the model is more likely to proactively flag uncertainty in its outputs, while substantially reducing unsupported claims. The company says this "more honest progress reporting" helps users build a more reliable chain of trust when collaborating with AI.

Beyond the core model upgrade, Anthropic has also rolled out three new features to enrich the Claude ecosystem. Analysts see the launch of Opus 4.8 as a continuation of Anthropic's "high quality, steady pace" upgrade strategy. The combination of unchanged pricing and enhanced performance could help solidify its premium positioning in an increasingly competitive AI model market. As the model becomes more autonomous, developers and enterprise users may reduce manual oversight costs—but balancing independent decision-making with safety and control remains a key focus going forward.