Glossary · RAG

Fixed-size Chunking

Also known as: Token-based chunking, Fixed-window chunking

Slice documents every N tokens with a small overlap. Fast and format-agnostic, but cuts through headings and mid-topic.

The simplest chunking strategy in RAG. Documents are split at fixed token boundaries (often 256 to 1024 tokens) with a small overlap so that context spanning a boundary is not fully lost. Runs at scale on any format but ignores meaning and structure.

In practice

The baseline every RAG interview compares against. Understanding why it degrades on structured docs is the setup for semantic and recursive chunking.

How it compares

vs. Semantic Chunking →

Fixed-size cuts every N tokens regardless of meaning; semantic chunks at topic boundaries by finding embedding-similarity drops.

vs. Recursive Chunking →

Fixed-size ignores the document; recursive splits by the document's own hierarchy of sections, paragraphs, and sentences.

Comparisons that include Fixed-size Chunking

Chunking strategies: fixed vs semantic vs recursiveFixed-size chunking is simple; semantic chunks at meaning boundaries; recursive respects the document's own structure. Recursive is the modern default for anything hierarchical (headings, code).

Related terms

Retrieval-Augmented Generation (RAG)

Fetch relevant docs at inference time and stuff them into the prompt so the model can answer from real data.

Chunking

Cutting documents into retrievable pieces before embedding them.

Context Window

The max number of tokens a model can attend to at once.

Hallucination

When a model confidently makes up something that isn't true.

AI System Design

End-to-end design of production LLM systems: ingestion, retrieval, serving, eval, monitoring.

HNSW (Hierarchical Navigable Small World)

A graph-based ANN index with log-time approximate search, excellent recall/latency trade-off.