Unverified50% confidenceFactExact time
Vector RAG has long hovered around 40% accuracy on relational queries.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
UnverifiedTraditional vector RAG generally achieves an accuracy rate of 68%–72% in complex multi-hop query scenarios.74% similarUnverifiedAWS's 2024 case study with Lettria showed that layering graph-based retrieval on top of vector RAG improved answer precision by up to 35% in finance, healthcare, aviation and legal use cases.73% similarUnverifiedMeta、NVIDIA等公司的研究表明RAG在知识密集型任务上的准确率比纯参数记忆提高20%-40%69% similarUnverified在多跳推理任务上 GraphRAG 的准确率比传统 RAG 高出 40% 以上，但索引构建时间增加 3-5 倍69% similarUnverifiedAccording to Logic42's Q3 2026 customer audit benchmark, pure vector search (RAG) has an average query latency of 180 milliseconds, 1200 tokens per query, a cost of $18 per 10,000 queries, and 41% relational accuracy.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1046421API
curl https://kongchang.com/api/v1/knowledge/claims/1046421MCP
get_claim(id=1046421)