这篇论文直击 AI 智能体重复计算同一文档的浪费痛点，做 LLM 推理优化或智能体基础设施的团队值得关注——它提出的 KV 缓存共享方案可能大幅降低推理成本，且已有实测数据支撑。
这篇论文指出当前 AI 智能体在处理相同文档时，每个智能体都会重复执行最耗计算的前缀填充（prefill）步骤，重建相同的键值（KV）缓存，造成巨大浪费。作者提出一个简单方案：让发布者预计算文档的 KV 缓存，其他智能体付费加载后跳过 prefill。实验表明，在 Qwen3-4B 模型上，复用比从头 prefill 节省 9-50 倍计算量，且输出 token 完全一致。但 KV 缓存体积大、难以压缩，直接传输成本高；作者建议在服务端托管（类似 prompt-caching），消除传输开销。以一个热门 3774 token 文档被 80M 智能体访问为例，复用计算成本仅约 3 万美元，而重新 prefill 需 150 万美元，节省 49.7 倍。论文还提出了智能体原生 prefill CDN 的框架，并指出无损 KV 压缩和跨方支付层是待解决的关键问题。
Can I Buy Your KV Cache?
Right now, across the world, AI agents are repeating the same absurd act: to read one document, they each recompute it from scratch. Every agent re-runs prefill, the most compute-intensive step a large model takes, over identical text, only to rebuild a key-value (KV) cache identical to the one the agent before it just built. The same answer, computed a million times. We make a proposal that is almost offensively simple: compute it once. Let a publisher precompute a document's KV cache, and let every other agent buy the right to load it and skip prefill. It works, and it is token-exact: loading a precomputed KV and continuing matches prefilling from scratch (24/24 greedy tokens, and at the logits level), with no accuracy cost. On Qwen3-4B, reuse is 9-50x cheaper in compute than prefill, and the gap widens with length (prefill's attention scales with L^2), so a single reuse already pays it back. Then the part that matters: where the KV lives. Shipping it fails, because KV is nearly incompressible, so per-load egress costs more than the prefill it saves. Hosting it provider-side, exactly as production prompt-caching works, removes egress entirely. The size of the prize is set by our measured compute saving: serving one hot 3774-token document to 80M agents costs ~$1.5M to re-prefill but only ~$0.03M of reuse compute (49.7x less). The 0.1x cache-read tariff APIs charge passes a 10x discount to users while sitting inside this measured envelope, so the 10x is a floor that the measured ~50x compute saving clears, and the gap to the physical ~50x is provider margin: millions of dollars per popular document. We frame the resulting agent-native prefill CDN and leave lossless KV compression and a cross-party payment layer as the open problems.