Anthropic has unveiled its latest flagship model, Claude Opus 5, which demonstrates impressive AI capabilities by achieving near-Fable 5 performance while costing half as much per token. According to The Decoder, Claude Opus 5 scored 30.2 percent on the ARC-AGI-3 benchmark, marking a significant advancement in artificial general intelligence metrics.
This new model's score on ARC-AGI-3 is nearly four times higher than that of GPT-5.6 Sol, highlighting its superior performance in coding and knowledge work. The Decoder emphasized that Claude Opus 5 maintains top scores in these areas while operating at a substantially reduced token price compared to Fable 5.
For Japanese markets, where AI-driven technologies are increasingly integrated into trading algorithms and data analysis, Claude Opus 5’s improved efficiency and cost-effectiveness could influence the adoption of advanced AI models in financial services and tech sectors.