人工智能公司Anthropic宣布推出一套新的报告工具，旨在帮助企业客户更清晰地了解其AI代币支出的具体价值和投资回报率。1 该工具允许经理查看团队工作产出、成本及工作流程瓶颈。1 此外，该工具利用内部分析和员工调查，展示团队工作范围、产出、成本及流程断点，并提供改进建议。1 针对数据隐私问题，Anthropic承诺默认不使用这些数据训练模型。1
这一发布背景是行业分析显示，尽管大多数财富500强企业尚未从AI部署中获得显著收益，但初创企业和特定领域对Claude等编码工具的使用量和支出正在快速增长。1 根据Jellyfish报告，基于Claude等工具提出的代码变更占审查总量的近三分之一；自年初以来，相关工具支出增长了10倍，约占工程总成本的2%。1 Jellyfish首席科学家Nicholas Arcolano指出，CFO们未预料到如此快速的支出增长，强调优化而非单纯投入的重要性。1
在商业表现方面，Anthropic在第二季度录得营业利润，净收入留存率达到900%，即同一客户群2026年Q3支出比2025年Q3增长九倍。1 虽然OpenAI的Astra模型被视为潜在威胁，但并未减缓Anthropic的企业采用率。1 Jellyfish AI产品负责人Nikolas Albarran表示，由于前期工作流标准化投入巨大，企业切换工具的转换成本高昂。1
Anthropic has launched a new suite of reporting tools designed to help enterprise clients understand the specific value and return on investment of their AI token spending. 1 The tools utilize internal analytics and employee surveys to display team work output, costs, workflow bottlenecks, and improvement suggestions. 1 Anthropic has committed that this data will not be used for model training by default. 1
The company recorded an operating profit in the second quarter and achieved a net revenue retention rate of 900%. 1 This rate indicates that spending from the same customer group is expected to increase ninefold between the third quarter of 2025 and the third quarter of 2026. 1 Industry analysis shows that while most Fortune 500 companies have not yet seen significant returns from AI deployment, usage and spending on coding tools like Claude are growing rapidly among startups and specific sectors. 1 A report by Jellyfish states that code changes proposed using tools such as Claude account for nearly one-third of all reviewed code, with related tool spending increasing tenfold since the start of the year and now comprising approximately 2% of total engineering costs. 1
Nicholas Arcolano, Chief Scientist at Jellyfish, noted that CFOs did not anticipate such rapid growth in spending, emphasizing the importance of optimization rather than simple investment. 1 Nikolas Albarran, Head of AI Products at Jellyfish, indicated that switching costs remain high for enterprises due to the substantial effort required to standardize workflows prior to adoption. 1 Although OpenAI’s Astra model is considered a potential threat, it has not slowed Anthropic’s enterprise adoption rates. 1
评论
还没有评论，欢迎留下第一条。