Anthropic has introduced its latest AI models, Claude Fable 5.1 and Mythos 5.1, marking the company’s most advanced releases to date. According to The Decoder, these models deliver significant improvements in both capability and efficiency.
Claude Fable 5.1 notably doubles the score of its predecessor on the Terminal-Bench-Science benchmark and achieves over a 30 percent boost in agentic coding performance. Additionally, operating costs for long, autonomous runs with multiple tool calls have been reduced by up to 45 percent, making the models more cost-effective for extensive usage.
For Japanese markets, where AI-driven automation and coding tools are increasingly integrated into financial services and technology sectors, these advancements could accelerate adoption and innovation, potentially impacting FX, crypto, and equities trading platforms.