Search method comparison

GraphRAG offers three distinct search methods, each optimized for different types of queries. This guide helps you understand when to use each method and how they compare.

Search methods overview

Global search

Map-reduce over community reportsBest for high-level, dataset-wide questions and thematic analysis.

Local search

Entity-centric retrievalBest for specific entity queries and detailed fact retrieval.

DRIFT search

Iterative graph traversalBest for complex multi-hop reasoning and exploratory analysis.

Quick comparison table

Aspect	Global Search	Local Search	DRIFT Search
Question Type	Broad, thematic	Specific, factual	Complex, multi-hop
Data Source	Community reports	Entities + text + reports	Dynamic graph traversal
Context Building	Fixed (all reports at level)	Semantic retrieval	Iterative expansion
LLM Calls	Many (map-reduce)	Single	Multiple (iterative)
Cost	High	Low-Medium	Medium-High
Response Time	Slow (2-10s)	Fast (<2s)	Medium (2-5s)
Token Usage	High	Low-Medium	Medium-High
Coverage	Entire dataset	Focused	Adaptive
Depth	Summarized	Detailed	Multi-level

When to use each method

Global search

Ideal for
Strengths
Limitations
Example

“What are the main themes in this dataset?”
“What are the top trends across all documents?”
“Summarize the key findings”
“What are the most significant events?”
“How do different topics relate globally?”

graphrag query \
  "What are the most important themes in this story?" \
  --method global

Local search

Ideal for
Strengths
Limitations
Example

“Who is Dr. Jordan Hayes?”
“What is the relationship between X and Y?”
“What are the properties of this entity?”
“When did this specific event occur?”
“What does the document say about X?”

graphrag query \
  "Who is Scrooge and what are his relationships?" \
  --method local

DRIFT search

Ideal for
Strengths
Limitations
Example

“How do organizations influence these events?”
“What chain of events connects A to B?”
“What patterns emerge from analyzing X, Y, and Z together?”
“How are these entities indirectly connected?”
“What are the multi-step dependencies?”

# Currently via API/notebooks only
result = await drift_search.search(
    "How do the different factions interact through intermediaries?"
)

Side-by-side examples

Here’s how each method handles the same dataset but different query types:

Dataset: “Operation Dulce” (sci-fi narrative)

Query 1: Broad thematic question

Question: “What are the main themes in this story?”

Global Search (Best)
Local Search (Suboptimal)
DRIFT Search (Moderate)

Response: Analyzes all community reports to identify overarching themes like government secrecy, alien contact, scientific ethics, and human cooperation.Why it works: Synthesizes information across the entire narrative structure.Cost: ~15,000 tokens

Query 2: Specific entity question

Question: “Who is Agent Mercer and what is their role?”

Local Search (Best)
Global Search (Suboptimal)
DRIFT Search (Good)

Response: Retrieves Agent Mercer entity, related relationships, and relevant text chunks providing detailed information about their role, background, and actions.Why it works: Direct entity retrieval with supporting evidence.Cost: ~2,500 tokens

Query 3: Multi-hop reasoning

Question: “How do the government organizations indirectly influence the scientific research at the base?”

DRIFT Search (Best)
Local Search (Limited)
Global Search (Moderate)

Response: Traces connections from government entities through intermediary actors to research activities, revealing multi-step influence patterns.Why it works: Designed for graph traversal and multi-hop reasoning.Cost: ~7,000 tokens

Query 4: Relationship query

Question: “What is the relationship between Dr. Hayes and the Dulce facility?”

Local Search (Best)
Global Search (Limited)
DRIFT Search (Good)

Response: Retrieves both entities and their connecting relationships with supporting text evidence.Why it works: Optimized for entity relationship queries.Cost: ~2,800 tokens

Decision flowchart

Is your question about a specific entity or relationship?

YES → Use Local SearchExamples:

“Who is X?”
“What is X’s relationship to Y?”
“What are X’s properties?”

NO → Continue to next step

Does your question require understanding the entire dataset?

YES → Use Global SearchExamples:

“What are the main themes?”
“What are the key trends?”
“Summarize this dataset”

NO → Continue to next step

Does your question involve multi-hop reasoning or complex connections?

YES → Use DRIFT SearchExamples:

“How does X indirectly affect Y?”
“What chain of events connects A to B?”
“What patterns involve X, Y, and Z?”

NO → Try Local Search first, then escalate if needed

Hybrid approaches

You can combine search methods for comprehensive analysis:

Sequential querying

Start with global search

Get high-level themes and important entities:

graphrag query "What are the main entities in this dataset?" --method global

Follow up with local search

Deep dive into specific entities identified:

graphrag query "Tell me more about [entity from global results]" --method local

Use DRIFT for connections

Explore relationships between key entities:

result = await drift_search.search(
    "How do [entity1] and [entity2] influence each other?"
)

Validation strategy

# Use multiple methods to validate findings

# Global: Identify main themes
global_result = await global_search.search(
    "What are the key themes?"
)

# Local: Verify themes with specific evidence
local_result = await local_search.search(
    "What evidence supports the theme of [X]?"
)

# DRIFT: Explore how themes connect
drift_result = await drift_search.search(
    "How do themes [X] and [Y] relate through entities?"
)

Performance benchmarks

Based on typical “Operation Dulce” dataset queries:

Metric	Global	Local	DRIFT
Avg Response Time	8.5s	1.2s	3.8s
Avg Tokens (Prompt)	11,500	2,800	6,200
Avg Tokens (Output)	800	400	600
Avg Cost (GPT-4)	$0.12	$0.03	$0.07
Parallelizable	Yes	No	Partially

Actual performance varies based on dataset size, query complexity, and configuration parameters.

Configuration comparison

Global Search
Local Search
DRIFT Search

# Key parameters
context_builder_params = {
    "max_tokens": 12_000,
    "community_level": 2,  # 0=coarse, 3+=fine
    "use_community_summary": False,  # Full vs summary
}

map_llm_params = {
    "max_tokens": 1000,
    "temperature": 0.0,
}

reduce_llm_params = {
    "max_tokens": 2000,
    "temperature": 0.0,
}

# Key parameters
local_context_params = {
    "text_unit_prop": 0.5,  # Text chunk allocation
    "community_prop": 0.1,  # Community report allocation
    "top_k_mapped_entities": 10,  # Entity retrieval count
    "top_k_relationships": 10,  # Relationship count
    "max_tokens": 12_000,
}

model_params = {
    "max_tokens": 2_000,
    "temperature": 0.0,
}

# Key parameters
drift_params = DRIFTSearchConfig(
    primer_folds=1,  # Initial retrieval rounds
    drift_k_followups=3,  # Expansions per iteration
    n_depth=3,  # Max traversal depth
)

Cost optimization strategies

Use local search first

Start with low-cost local search; escalate only if needed

Adjust community level

Use level 1 for global search instead of level 2 to reduce tokens

Tune DRIFT conservatively

Keep n_depth=2 and drift_k_followups=2 for most queries

Use summaries

Set use_community_summary=True in global search

Common pitfalls

Using global search for specific entities

Problem: Expensive and provides less detail than local search.Solution: Use local search for entity-specific queries.

Using local search for broad themes

Problem: May miss important information not connected to retrieved entities.Solution: Use global search for dataset-wide questions.

Over-parameterizing DRIFT

Problem: Setting n_depth=5 and drift_k_followups=10 wastes tokens.Solution: Start with default parameters and increase only if needed.

Not checking context data

Problem: Accepting answers without verifying supporting evidence.Solution: Always inspect result.context_data to see what was used.

Choosing based on use case

Research & analysis
Q&A systems
Investigation
Content summarization

Primary: Global search for themes and patternsSecondary: Local search for validating specific claimsTertiary: DRIFT search for exploring connections between findings

Next steps

Global search

Deep dive into global search

Local search

Master local search techniques

DRIFT search

Learn DRIFT search methods

Query overview

Complete query documentation

Tutorials

Notebooks

Use Cases

Search method comparison

Search methods overview

Global search

Local search

DRIFT search

Quick comparison table

When to use each method

Global search

Local search

DRIFT search

Side-by-side examples

Dataset: “Operation Dulce” (sci-fi narrative)

Decision flowchart

Hybrid approaches

Sequential querying

Validation strategy

Performance benchmarks

Configuration comparison

Cost optimization strategies

Use local search first

Adjust community level

Tune DRIFT conservatively

Use summaries

Common pitfalls

Choosing based on use case

Next steps

Global search

Local search

DRIFT search

Query overview

Build docs developers (and LLMs) love

Tutorials

Notebooks

Use Cases

​Search methods overview

Global search

Local search

DRIFT search

​Quick comparison table

​When to use each method

​Global search

​Local search

​DRIFT search

​Side-by-side examples

​Dataset: “Operation Dulce” (sci-fi narrative)

​Decision flowchart

​Hybrid approaches

​Sequential querying

​Validation strategy

​Performance benchmarks

​Configuration comparison

​Cost optimization strategies

Use local search first

Adjust community level

Tune DRIFT conservatively

Use summaries

​Common pitfalls

​Choosing based on use case

​Next steps

Global search

Local search

DRIFT search

Query overview

Build docs developers (and LLMs) love

Search methods overview

Quick comparison table

When to use each method

Global search

Local search

DRIFT search

Side-by-side examples

Dataset: “Operation Dulce” (sci-fi narrative)

Decision flowchart

Hybrid approaches

Sequential querying

Validation strategy

Performance benchmarks

Configuration comparison

Cost optimization strategies

Common pitfalls

Choosing based on use case

Next steps