Artificial intelligence startup Anthropic on Friday unveiled Claude Opus 5, its newest flagship AI model, saying it delivers stronger performance than Claude Fable 5 while costing half as much, as businesses increasingly demand lower-cost AI tools with clearer returns on investment.<\/p>\n
According to Anthropic, Opus 5 outperforms its predecessor on coding and knowledge-work benchmarks and is designed for everyday enterprise use, News.Az<\/a> reports, citing CNBC<\/a>.<\/p>\n The company said the model will cost $5 per million input tokens and $25 per million output tokens, compared with the higher pricing of Fable 5.<\/p>\n While Anthropic described Opus 5 as its best-performing and most cost-effective model to date, it said the model is not state-of-the-art for high-risk, dual-use applications, including offensive cybersecurity.<\/p>\n The launch comes as Anthropic and rival OpenAI face increasing pressure from enterprise customers looking to control AI spending and justify investments with measurable business value.<\/p>\n \u201cEnterprises, in our feedback and with our customer base, are looking for value,\u201d Dianne Penn, Anthropic’s head of product management for research, told CNBC. \u201cIf it’s a cheaper model or a cheaper offering, but it’s not accomplishing a similar level of quality, it’s actually not useful.\u201d<\/p>\n Anthropic is also competing with a growing number of lower-cost AI offerings from companies including Microsoft, Amazon, Google, and several Chinese AI startups.<\/p>\n The company has faced heightened scrutiny since introducing Claude Mythos Preview in April, a model it said demonstrated advanced cybersecurity capabilities, including identifying and exploiting software vulnerabilities.<\/p>\n In June, Anthropic<\/a> released Mythos 5 alongside Fable 5, describing the latter as safe for general use and its most capable publicly available AI model at the time.<\/p>\n Both models were temporarily taken offline days after launch to comply with a government export control directive citing national security concerns. Anthropic restored access roughly two weeks later after negotiations with government officials.<\/p>\n Penn said Anthropic continues to work closely with government agencies during pre-deployment testing of new models, including Opus 5. According to the company, Mythos 5 still exceeds Opus 5 in offensive cybersecurity and biology research capabilities.<\/p>\n