1
中兴推出“星云 · 晴天全屋套装 BE3600 PoE 版”全屋组网 Wi-Fi 7 路由产品，“1 母 2 子”售 949 元
IT之家 8 月 27 日消息，中兴现已在京东上架“星云 · 晴天全屋套装 BE3600 PoE 版”全屋组网 Wi-Fi 7 路由产品，采用“1 母 2 子”搭配，定价为 1049 元， 首发价 949 元 。用户购买单独子路由价格为 309 元。 京东 中兴（ZTE）晴天 PoE 全屋路由 BE…
2
TailSieve: Partial-Rollout-Guided Tail Routing for LLM Rollouts
LLM推理新招：用部分展开引导尾部路由，直击长尾生成瓶颈，性能提升值得关注。
arXiv:2608.22788v1 Announce Type: new Abstract: Large-scale rollouts have become a core component of modern LLM systems, spanning reinforcement learni…
3
Most of the LLM routing gap is task type
揭秘LLM路由差距的根源：任务类型比模型选择更关键，实测14模型两次运行结果并不完全一致。
arXiv:2608.23023v1 Announce Type: new Abstract: An LLM router picks which model should answer each query. The appeal is that models fail on different …
4
Self host sheets, docs, drive, email
自托管协作套件，洋葱路由加三服务隔离架构，实时协作延迟出乎意料地可用，多平台可部署。
Selfhost a full suite of collaboration tools powered by onion v3 routing and a isolated three dedicated service architecture. Surprisingly usable late…
5
华硕宣布 3 款 Wi-Fi 8 无线路由器“很快上市”
IT之家 8 月 25 日消息，华硕 (ASUS) 今日表示，其 3 款支持新一代 WLAN 规范 Wi-Fi 8 (802.11bn) 的无线路由器将 “很快上市” 。 华硕的 Wi-Fi 8 初期阵容包括两款电竞路由器 ROG Rapture GT-BN98 Pro、ROG Rapture GT…
6
Ramp launches its own AI model router, called Router
Ramp自研AI模型路由器，整合多家大模型，免费试用至年底，企业费用管理迎来新入口。
Ramp has launched its own AI model routing service, dubbed Router, that lets users and companies use and switch between various large language models …
7
支付巨头 Stripe 确认将收购 AI 路由平台 OpenRouter，消息称交易价格达 75 亿美元
支付巨头 Stripe 75 亿美元收购 AI 路由平台 OpenRouter，AI 与支付生态融合再提速
IT之家 8 月 20 日消息，在线支付服务商 Stripe 当地时间周三确认，公司将收购 OpenRouter。虽然 Stripe 没有公布这笔交易的具体金额，但知情人士告诉《纽约时报》，交易价格达到 75 亿美元 （IT之家注：现汇率约合 505.96 亿元人民币） 。 这一价格远高于 Open…
8
LLMs Can Predict Failure Risk, But Struggle to Predict Which Collaboration Protocol Pays Off: Cost-Aware Protocol Routing Across Reasoning Tasks
大模型能预判推理失败，却难选协作协议？这篇论文提出成本感知路由，高效平衡推理性能与开销。
arXiv:2608.14927v1 Announce Type: new Abstract: Multi-agent large language model (LLM) systems can improve reasoning by spending more computation, but…
9
Launch HN: Speko (YC S26) – OpenRouter for Voice AI
语音AI界的OpenRouter，按语言与场景智能路由多模型，不再迷信厂商榜。
Hi HN! I'm Bek, founder of Speko, a platform that finds an optimal combination of speech-to-text, LLM, and text-to-speech models, given your constrain…
10
Not All Tokens Are Equal: Inflation-Aware Routing for Agentic LLM Systems
大模型重试机制隐藏巨额成本，这份研究用“Token膨胀率”重新定义路由策略，为Agent系统省钱指明新方向
arXiv:2608.13571v1 Announce Type: cross Abstract: When a language model fails to answer a query on the first attempt, an agentic system retries, consu…
11
请立即更新：爱快 iKuai 路由器系统近期遭受持续攻击
爱快iKuai路由器遭持续攻击，DNS功能受影响且可能断网，官方建议立即升级至4.0.307及以上版本防患未然。
IT之家 8 月 17 日消息，爱快安全团队前天在官方论坛发文称，爱快 iKuai 路由器系统近期遭受持续攻击，虽然攻击具体源头目前仍不明确，但官方已采取了反制措施。 IT之家从官方公告获悉，本次攻击主要影响 DNS 功能，还针对部分路由强制断网。通过对异常行为的样本分析，官方已定位到部分受影响的客…
12
What Drives LLM Self-Reflection? A Controlled Ablation of Uncertainty Routing in Armed Conflict Forecasting
用消融实验揭开LLM自我反思的机制，看不确定性路由如何影响冲突预测准确率。
arXiv:2608.12322v1 Announce Type: cross Abstract: Self-reflection is widely assumed to improve LLM reasoning, yet which component drives the gain rema…
13
LLM Router: Rethinking Routing with Prefill Activations
用prefill激活值重思LLM路由选择，为多模型推理调度开辟高效新路径。
arXiv:2603.20895v3 Announce Type: replace-cross Abstract: Existing routers rely on semantic query features or handcrafted features, which often fail t…
14
Leanroute Is Live One AI Gateway for Models and Tools
AI网关进化新方向，统一调度大模型与MCP工具流量，为AI应用提供一体化接入层。
Hi all, We built Leanroute - it's One AI Gateway for LLM and Tools. Market is flooded with LLM Gateways, so I wanted to build an AI Gateway for LLM an…
15
RouteGuard: Certifying Routing Gain in LLM Multi-Agent Systems When Complementarity Is Not Enough
多智能体LLM路由增益难预判，RouteGuard认证框架揭示只看AUC和互补性远远不够。
arXiv:2608.07583v1 Announce Type: cross Abstract: Multi-agent LLM systems route among model-backed advisors, yet a deployer rarely knows before shippi…
16
LLM Reasoning for Subjective Tasks: Failure Modes, Mitigation, and Dynamic Reasoning Routing
LLM推理并非万能，主观任务中会失灵？这篇论文剖析失败模式并给出动态路由方案。
arXiv:2608.08889v1 Announce Type: new Abstract: Recommendation systems thrive on personalization, where ''correctness'' is rarely a binary truth but a…
17
Show HN: Pacific Slate: a self-hosted, model-agnostic multi-agent AI assistant
自托管多代理AI助手，模型无关且能学习偏好，把私有数据变成你的专属智能管家
I didn't want to buy a standalone computer or repurpose a laptop to run constantly so I could maintain a system to sync my LLMs, so I built this. It's…
18
Databricks drove down AI coding spend 70%
Databricks 实战分享：混合模型路由让 AI 编码成本直降 70%，大厂省钱经验值得抄作业。
Article URL: https://www.databricks.com/blog/managing-ai-coding-costs-scale Comments URL: https://news.ycombinator.com/item?id=49214468 Points: 142 # …
19
Building an AI Chat for My Blog: Gemini, Cloud Functions, $5-15/Month
为博客搭建AI聊天助手，用Gemini和Cloud Functions实现低成本智能问答，月费仅5-15美元。
Article URL: https://emergencemachine.com/building-an-ai-chat-for-my-blog/ Comments URL: https://news.ycombinator.com/item?id=49212021 Points: 1 # Com…
20
MACRO: Markov Chain Routing of Transformer Layers
用马尔可夫链给Transformer层做路由，动态跳过冗余层，推理效率有望大幅提升。
arXiv:2608.05872v1 Announce Type: cross Abstract: Standard Large Language Models (LLMs) execute layers sequentially. Dynamic layer routing, i.e. searc…