GraphRAG vs. plain RAG: the honest scorecard

Key points

• GraphRAG won 72 to 83 percent of head-to-head comprehensiveness comparisons against standard RAG in Microsoft's own testing.
• Graph-guided retrieval raised average recall scores from 73.4 percent to 87.8 percent on multi-step question benchmarks.
• Building the knowledge graph cost roughly $48 per corpus with GPT-4o, far above a basic vector index.
• On simple, single-fact lookups, plain text chunks matched or beat graph methods in a 2025 Michigan State and Meta study.
• Microsoft's lower-cost follow-up approach, LazyGraphRAG, cut the indexing price to about 0.1 percent of the original, admitting the full build is too expensive for many teams.

Search the right four words inside a document and you get an instant, useful answer. Ask what the recurring complaints were across two years of customer emails, and the same system shrugs.

That gap is the problem GraphRAG was built to fix.

What is GraphRAG, and why does it exist?

Standard RAG, which stands for retrieval-augmented generation, works like a well-organised filing cabinet. You ask a question, the software finds the most relevant slips of paper (called chunks), and hands them to an AI language model to write an answer. Fast, cheap, good for direct questions.

But a filing cabinet cannot connect the dots. If the answer lives across a dozen files, same customer complaint, different months, different departments, no individual slip of paper contains it. The model never sees the link.

GraphRAG takes a different approach. Before anyone asks a question, a large language model (the AI technology behind tools like ChatGPT) reads the entire document collection and maps out all the named things inside it: people, companies, dates, events, policies. It notes how those things relate to each other, like a spider diagram on a whiteboard. That diagram is the knowledge graph.

The software then groups the diagram into clusters of related topics and writes a plain-English summary for each cluster. When a question arrives, those summaries do the heavy lifting instead of raw document snippets.

When does the graph actually win?

The honest answer, reported by VentureBeat after reviewing five studies including Microsoft's original research, is: it depends entirely on the question type.

Question type	Plain chunks score	Graph score	Winner
Simple fact lookup	60.9	60.1	Tie
Multi-step reasoning	42.9	53.4	Graph +10 pts
Whole-corpus summary	51.3	64.4	Graph +13 pts
Global comprehensiveness	baseline	+72 to 83% win rate	Graph

For questions that require joining scattered facts or summarising a large body of material, the graph wins clearly. For a quick factual lookup, the extra machinery adds cost without adding accuracy.

Recall scores (the share of truly relevant passages the system actually finds) also jumped with graph-guided retrieval: from 73.4 percent up to 87.8 percent on average across standard multi-step question tests.

What does it cost, and is the scoring fair?

Two problems deserve plain-language honesty.

Building the graph is not free. One analysis put the indexing cost at around $48 per corpus using OpenAI's GPT-4o model. Microsoft's own cheaper follow-up, called LazyGraphRAG, defers most of that work and cuts the price to roughly 0.1 percent of the original figure. That gap tells you the full build is impractical for many real-world teams.

Many of the performance comparisons also used another AI model as the judge, which introduces bias. Auditors found that simply swapping the order in which two answers appear can shift the winner by more than 30 percentage points. One method's reported 66.7 percent win rate fell to about 39 percent after corrections.

The large recall and accuracy gains still hold up. Narrow comprehensiveness margins deserve more scepticism.

What should a team actually do?

The practical answer from the research is a hybrid: route simple fact questions to plain chunks, and send complex, multi-part questions to the graph. Combining both methods consistently beat either one alone in controlled tests.

You do not need to pick sides. You need a system that picks for you.