The commercialization of LLM applications is the next frontier in online advertising, with LLM-native advertising emerging as a promising paradigm by integrating ads into LLM-generated content. However, classic mechanisms are no longer applicable in this setting where the auction object is shifted from discrete ad slots to distributions over LLM outputs, and existing methods are impractical in industrial scenarios due to ignored externalities or high inference costs. To address these issues, we propose LLM-Auction, the first learning-based generative auction mechanism that integrates auction and generation. By formulating the allocation as preference alignment between LLM outputs and a mechanism objective that balances advertisers' value and user experience, we optimize the LLMs to inherently model allocation externalities without extra inference cost. Theoretically, we identify the allocation monotonicity and continuity of LLM-Auction, and prove that a simple first-price payment rule exhibits favorable incentive properties. Furthermore, we build an LLM-as-a-judge simulation environment for quantitative evaluation, and experiments demonstrate that LLM-Auction achieves the state-of-the-art allocation efficiency while satisfying key mechanism properties.

翻译：大型语言模型（LLM）应用的商业化是网络广告的下一个前沿领域，其中LLM原生广告通过将广告融入LLM生成内容，成为一种极具前景的范式。然而，当拍卖对象从离散广告位转变为LLM输出的分布时，经典机制不再适用；同时，由于忽略外部性或推理成本过高，现有方法在工业场景中缺乏实用性。为解决这些问题，我们提出LLM-Auction——首个基于学习的生成式拍卖机制，该机制整合了拍卖过程与内容生成。通过将分配问题形式化为LLM输出与机制目标（平衡广告主价值与用户体验）之间的偏好对齐，我们优化LLM以内在建模分配外部性，且无需额外推理成本。理论上，我们证明了LLM-Auction满足分配单调性与连续性，并证明简单的首价支付规则具有良好激励性质。此外，我们构建了LLM-as-a-Judge模拟环境以进行定量评估，实验表明，LLM-Auction在满足关键机制性质的同时，实现了最优的分配效率。