Anthropic releases Claude Opus 4.8 with improved performance

Anthropic has launched Claude Opus 4.8, the latest upgrade to its flagship AI model, offering enhanced performance in coding, reasoning, and honesty while maintaining its previous pricing. The model is available globally as of May 28, 2026.

The new version builds upon the Opus 4.7 model released just six weeks earlier, marking a faster upgrade cycle for the company. Anthropic states Opus 4.8 demonstrates measurable improvements across several benchmarks, including agentic coding, multidisciplinary reasoning, and knowledge work. Early testers report the model exhibits sharper judgment and greater reliability in long-running, autonomous tasks.

A key advancement highlighted by the company is the model's improved honesty and self-awareness. Anthropic's evaluations indicate Opus 4.8 is approximately four times less likely than its predecessor to allow flaws in code it generates to pass without comment. The model is also described as more likely to flag uncertainties in its work and less likely to make unsupported claims. Alignment assessments show the model reaches new highs in prosocial traits and exhibits substantially lower rates of misaligned behavior compared to Opus 4.7.