看 Jev + Mercury 2.5 怎么用超低成本，在 WebMCP 测试里把 49 个任务全做完，比 GPT-6 Astra 便宜 100 多倍。
Jev + Mercury 2.5 在 WebMCP 基准测试中完成了全部 49 项任务。该模型成本比 GPT-6 Astra 配合代码执行低约 112 倍。其成本也比 Astra 使用截图操作浏览器低约 245 倍。
1/ Jev + Mercury 2.5 几乎“打穿”了 WebMCP 基准测试： 49/49 项任务全部完成。 模型成本比 GPT-6 Astra 配合代码执行低约 112 倍，比 Astra 使用截图操作浏览器低约 245 倍。👇 💬 3 🔄 0 ❤️ 2 👀 770 📊 3 ⚡
The result shows cheap small-model passes can beat LLM rerankers when split into selection and ranking stages, cutting model-call latency without losing recall on multi-hop questions. Teams building Graph RAG on Milvus can copy the two-stage pattern directly from the Zilliz repo. Compared with LLM-based reranking, the trade-off is a measurable recall gain for under a second of extra call time.
OpenBMB 聊了聊他们的 MiniCPM-OCR 和 Jev：不光提取文档信息，还会判断哪些结果需要再核一遍，做文档处理可以看看。
The interesting shift is that retrieval becomes a tree-traversal decision problem instead of an embedding-similarity problem, so latency scales with tree depth rather than corpus size, with billion-paragraph corpora theoretically needing around five decisions. Teams paying for a reranker stage, or struggling with weak first-stage recall on help centres and document corpora, can test the tree-plus-decision-model setup; on NQ a dense retrieval stage can optionally be mixed in to recover over-pruned evidence. It competes with Qwen3-Reranker-8B pipelines at lower cost.
一篇把判别式决策模型搬到卫星通信上的论文，单次前向 90 毫秒出决策，准确率还比生成式基线高出一大截，做边缘部署的可以看看。
Headline accuracy is the wrong metric when choosing a RAG claim checker, because a verifier can look accurate while waving through a large share of unsupported claims. Teams building retrieval-backed products should benchmark false-approval rates alongside cost before wiring in a cheap verifier, and cascade designs may not save more than random routing. The finding applies to any RAG stack where a wrong approval is costlier than a wrong rejection.
