Sentence Window Retrieval
A RAG technique that indexes individual sentences for precise matching but retrieves surrounding sentences (a window) for context. Provides fine-grained retrieval precision while maintaining adequate context for LLM generation.
About this tool
Overview
Sentence Window Retrieval indexes individual sentences but retrieves a window of surrounding sentences. This balances precision (finding exact relevant sentences) with context (providing enough information for the LLM).
How It Works
- Index: Each sentence separately with metadata about position
- Search: Find most relevant sentences
- Expand: Retrieve N sentences before and after
- Return: Combined window as context
Configuration
from llama_index import SentenceWindowNodeParser
node_parser = SentenceWindowNodeParser.from_defaults(
window_size=3, # 3 sentences before and after
window_metadata_key="window",
original_text_metadata_key="original_text",
)
Benefits
- Fine-grained matching (sentence-level precision)
- Flexible context window
- Less redundancy than overlapping chunks
- Natural sentence boundaries
Use Cases
- Question answering requiring specific facts
- Citation and attribution
- Legal/medical documents
- When precision matters more than broad context
Pricing
Implementation-dependent (LlamaIndex feature).
Surveys
Loading more......
Information
Websitedocs.llamaindex.ai
PublishedMar 15, 2026
Categories
Tags
Similar Products
6 result(s)