Contextual Retrieval
Contextual Retrieval
Why it matters
Prepending 50–100 tokens of LLM-generated context to each chunk cut the retrieval failure rate by 35%; by 49% when combined with BM25, and by 67% with reranking added
Source — Primary sources first; secondary reports are labelled “Press”
Officialanthropic.comhttps://www.anthropic.com/news/contextual-retrievalAdded in Week 35, 2026 · See that week