多智能体协作面临延迟挑战？这篇论文教你用学习驱动编排，显著提升系统响应效率。
arXiv:2601.10560v2 Announce Type: replace-cross Abstract: Multi-agent systems (MAS) coordinate multiple LLM-powered agents through structured workflow…
用叙事约束为 LLM 减负：token 消耗降低 79%，系统延迟从 3.4s 降到 1.3s，语义热力学带来全新优化视角。
Article URL: https://github.com/tauansloboda/semantic-thermodynamics Comments URL: https://news.ycombinator.com/item?id=49190710 Points: 2 # Comments:…
用WebCMD范式打造毫秒级Hacker News CLI，让AI代理绕过浏览器开销，延迟降95%，token省90%。
How we leveraged the WebCMD paradigm to slash web-navigation latency by 95%, reduce LLM token consumption by 90%, and establish a deterministic, schem…
大模型推理的“扁平延迟”困境，为何优化如此棘手？
Article URL: https://aidoses.substack.com/p/nothing-is-easy-when-youre-an-llm Comments URL: https://news.ycombinator.com/item?id=49125443 Points: 1 # …
TF-IDF vs 神经嵌入：LLMSlim 如何用微秒级延迟实现确定性提示压缩，架构决策值得一看。
Most prompt compression discussions focus on the happy path: you have a long RAG context, you trim it to 50% of tokens, and your API bill halves. What…
NVIDIA官方详解硬件感知的大模型设计，平衡吞吐量与延迟的Pareto前沿策略。
Article URL: https://developer.nvidia.com/blog/ai-model-co-design-hardware-friendly-llm-design/ Comments URL: https://news.ycombinator.com/item?id=488…
让LLM直接操作bash终端，轻量级快照平衡速度与安全，但潜在风险恐怖。
Recently, I was testing an autonomous AI agent in a local directory. I gave it a multi-step coding task and a terminal tool so it could run its own te…
语音代理中流式音频的延迟难题，从原理到开源解决方案，技术深度与实战兼备。
Blog: https://gokuljs.com/blogs/when-latency-becomes-audible code: https://github.com/gokuljs/GoSFU Comments URL: https://news.ycombinator.com/item?id…
业务走向全球，多区域架构如何兼顾性能与成本？看这篇就够了。
When an application grows geographically, it is logical to start serving it from a second region to improve latency and availability.
AI Agent任务中LLM调用链被阶梯式中断，3-8倍延迟竟源于调度器无视中间状态
arXiv:2605.00528v2 Announce Type: replace-cross Abstract: AI agents execute tens to hundreds of chained LLM calls per task, yet GPU schedulers treat e…
实时语音AI的传输选型关键：为何WebRTC在延迟与拥塞控制上完胜WebSockets，做语音Agent必读。
Article URL: https://livekit.com/blog/why-webrtc-beats-websockets-for-voice-ai-agents Comments URL: https://news.ycombinator.com/item?id=48616386 Poin…
客户端数据库能省掉频繁的服务器往返，让应用响应速度有质的提升，别再忽视这种模式了。
For the last five years, I've worked on web apps that follow the same pattern: build a backend, set up Postgres, wire up REST endpoints, fetch data on…
揭秘Google搜索自动补全系统如何承受每秒千万级请求与100毫秒硬性延迟
Section 1 — Problem Definition An autocomplete system predicts and suggests query completions as a user types, character by character. The goal is to …
探讨RAG系统中缓存复用安全性的新方法，提出Grounded Cache Routing机制，优化成本与延迟。
arXiv:2605.27494v1 Announce Type: cross Abstract: Modern retrieval-augmented generation(RAG) deployments increasingly rely on caching to reduce token …
针对长时LLM Agent的上下文溢出问题，提出并行压缩方法，减少数十秒推理阻塞。
arXiv:2605.23296v1 Announce Type: new Abstract: Long-horizon LLM agents accumulate growing conversation histories that eventually exceed the model's c…
实时语音AI的延迟生死线：从Go到Rust的迁移如何压缩毫秒级时延，守住250ms实时交互边界。
In building Vivik, an execution-grade telephony AI engine, we faced a brutal constraint: the human conversational loop. In psychoacoustics, a delay un…
无需微调模型，无需修改函数协议，AsyncFC 在纯执行层实现异步并发，让 LLM 解码与函数执行重叠，大幅降低端到端延迟。LLM 对未执行结果（symbolic futures）的推理能力被天然利用，开启模型-工具异步交互新范式。
arXiv:2605.15077v1 Announce Type: cross Abstract: Function calling, also known as tool use, is a core capability of modern LLM agents but is typically…