Anthropic has unveiled its new flagship AI model, Claude Opus 5, which claims to deliver performance levels comparable to Fable 5 at just half the token price. This development is particularly noteworthy as Opus 5 has achieved impressive scores in coding and knowledge work, positioning it as a formidable contender in the rapidly evolving AI market. On the ARC-AGI-3 benchmark, which assesses novel problem-solving capabilities, Opus 5 recorded a score of 30.2 percent, nearly quadrupling the performance of its closest rival, GPT-5.6 Sol. Such advancements highlight Anthropic's commitment to enhancing AI efficiency while reducing operational costs for users.
Source: The Decoder
