央广网上海9月12日消息（记者 殷雨婷）9月12日，蚂蚁数科在2025Inclusion·外滩大会论坛上宣布，将推出按"效果付费"的新型商业模式。该模式支持 企业客户根据大模型应用的实际效果（如业务增长或成本节省）来付费，而非传统的项目制或订阅制模式。 随着大模型从技术层向场景层渗透，AI已成为业务成果的加速器，"企业级AI服务的衡量标准，应从'为技术付费'向'为效果付费'转变，一些以结果为重 的产品，客户可以选择为可见的效果买单这种付费方式。"蚂蚁数科副总裁余滨表示，目前蚂蚁数科已经与部分战略级客户达成基于关键业务效果的合作模 式，将逐步向全行业推广。 蚂蚁数科副总裁余滨在2025外滩大会演讲。受访者供图 "按效付费"模式有望显著降低企业应用大模型的门槛和风险，结合蚂蚁数科的全栈企业级AI服务落地方案体系，尤其有助于中小企业更轻松地接入AI能 力，加速大模型在金融、能源、零售等千行百业的规模化落地。 余滨表示，蚂蚁数科基于行业实践，已经形成大模型落地产业的"ACE"（Align-Construct-Evaluate）三步方法论：首先是与客户进行价值定位，一同找 到可衡量的、最有价值的业务目标。第二步，根据业 ...

Core Insights - The introduction of a new "pay-for-performance" business model by Ant Group's Ant Financial, allowing enterprise clients to pay based on the actual results achieved from AI model applications, rather than traditional project-based or subscription models [1] - The shift from AI as a standalone technology tool to an accelerator of business outcomes, emphasizing the need for enterprise-level AI services to focus on measurable results [1] - The "pay-for-performance" model aims to lower the barriers and risks associated with large-scale AI adoption, particularly benefiting small and medium-sized enterprises by reducing upfront costs and infrastructure investments [1] Summary by Sections Business Model Transformation - Ant Financial announced a new business model that supports clients in paying based on the effectiveness of AI applications, marking a significant shift in the enterprise service market [1] - The model is designed to facilitate easier access to AI capabilities for businesses, especially in sectors like finance, energy, and retail [1] Challenges in AI Adoption - High initial costs, uncertain returns, and lack of implementation pathways are identified as major challenges for enterprises in scaling AI applications [1] - The new model is expected to mitigate these challenges by allowing businesses to invest in AI based on visible outcomes [1] Methodology for AI Implementation - Ant Financial has developed a three-step methodology called "ACE" (Align-Construct-Evaluate) for implementing large-scale AI solutions [2] - The first step involves value positioning with clients to identify measurable business goals, followed by constructing a comprehensive solution and implementation path, and finally ensuring business outcomes through continuous joint operations [2]