NEW

The wait is over. SurrealDB 3.0 is here!

Build intelligent knowledge graphs with SurrealDB

Q: How does SurrealDB model graph relationships?

SurrealDB uses record links and graph edges as first-class concepts. You create relationships with the RELATE statement, which produces typed edges that can carry their own properties. No separate graph database is needed.

Q: Can I traverse graphs with SurrealQL?

Yes. SurrealQL supports multi-hop graph traversals using arrow syntax. You can walk relationships in any direction, filter at each hop, and return nested structures — all in a single query without joins or subqueries.

Q: Can I combine graph queries with full-text and vector search?

Yes. Because SurrealDB is multi-model, you can mix graph traversals with full-text search, vector similarity, and document filters in the same query. This is especially powerful for knowledge graphs that need semantic retrieval alongside structural navigation.

Q: Does SurrealDB support recursive graph queries?

Yes. SurrealQL allows recursive traversals to explore graphs to an arbitrary depth. You can set depth limits, apply filters at each level, and collect results along the path for use cases like shortest-path analysis and dependency mapping.

Q: Do I need a fixed schema for my knowledge graph?

No. SurrealDB supports schemaless, schemafull, and mixed modes. You can start building your graph without defining a rigid schema upfront, then gradually add type constraints and validations as your model matures.

Graph-native multi-model database enabling rich semantics, real-time traversal, and reasoning - turning knowledge graphs into a programmable foundation for intelligent applications and AI agents.

Why knowledge graphs matter

Knowledge graphs transform scattered data points into a connected, machine-understandable network that powers deeper insights, stronger reasoning, and more reliable AI.

Connected data

Knowledge graphs model information as nodes and edges, creating clear relationships that make data easier to navigate and understand.

AI context and memory

They give AI and LLMs a reliable semantic layer for context and memory, grounding outputs in connected information for more consistent reasoning.

Multi-hop reasoning

Graphs enable multi-step traversals and pattern detection, supporting insights and queries that simple keyword or vector search can't provide.

Unified context

They combine structured data, documents, events, and embeddings into one connected system, giving applications richer context to work with.

THE CHALLENGE

The challenges with traditional graph databases

Legacy limits

Legacy graph databases were built for large data analysis, not for real-time updates, multi-model data, and the demands of modern AI workloads.

Fragmented data

Data gets scattered across multiple systems, making it hard to keep a clear and connected picture of your domain.

Operational overhead

Each hop between systems adds latency, ops overhead, and failure modes that complicate your architecture.

Complex stacks

Building AI or knowledge graph workloads requires several databases and pipelines, increasing latency and operational burden.

THE SOLUTION

A multi-model graph database for the AI era

Real-time graphs

SurrealDB supports continuous reads, writes, and relationship updates without cache fragility - ideal for fast-changing AI memory and agent workloads.

Native multi-model

Graphs, documents, vectors, events, and relational data work together seamlessly, enabling rich queries across all data types in one system.

Elastic scaling

A decoupled compute-and-storage design provides predictable horizontal scaling with no manual sharding or complex clustering.

Developer-friendly

SurrealQL combines familiar SQL-like syntax with graph and JSON querying, making graph and multi-model development fast and accessible.

CAPABILITIES

Model knowledge and relationships

Create entities and connect them with a typed edge containing metadata.

1-- Create entities (nodes)
2CREATE person:marie CONTENT { name: "Marie Curie", occupation: "Scientist" };
3CREATE award:nobel_prize CONTENT { category: "Chemistry", year: 1911 };
4
5-- Create a relationship (edge) with metadata
6RELATE person:marie->won->award:nobel_prize SET reason = "Discovery of Radium and Polonium";

Graph traversal: querying linked data

Traverse the graph to find all awards won by a person or all people who have won a specific category of award.

1-- Find all awards won by Marie Curie
2SELECT ->won->award.* FROM person:marie;
3
4-- Find all people who have won a Nobel Prize in Chemistry
5SELECT <-won<-person.* FROM award WHERE category = "Chemistry";

Combining graph with document data

Store rich document-style information on nodes and link them to other entities via relationships.

1-- Each node can store rich document data
2CREATE person:einstein CONTENT {
3	name: "Albert Einstein",
4	biography: {
5		birth: "1879-03-14",
6		nationality: "Swiss",
7		summary: "Developed the theory of relativity."
8	},
9	tags: ["physics", "relativity", "nobel laureate"]
10};
11
12-- Create a graph edge from Albert Einstein to the Nobel Prize
13RELATE person:einstein->won->award:nobel_prize SET year = 1921;
14
15-- Graph traversal with metadata filtering to find all people who won a Nobel Prize after 1910
16SELECT <-won<-person.* FROM award WHERE year > 1910;

SECURITY

Enterprise-ready by design

Security, scalability, and reliability built directly into the core of SurrealDB.

Built-in security

Fine-grained access control, authentication, and permissions are enforced directly at the database level to protect data across teams and applications.

Scale with confidence

Designed to scale horizontally and handle high-throughput, real-time workloads without sacrificing performance or reliability.

Flexible deployment

Run SurrealDB on-premises, in the cloud, or at the edge - giving enterprises full control over data residency and infrastructure.

Operational reliability

Strong consistency, fault tolerance, and predictable behaviour ensure SurrealDB can support mission-critical production systems.

Real-world scenarios

From fraud prevention to supply chain visibility, knowledge graphs power critical business applications

Fraud prevention

Link card transactions, devices, IPs, and identity documents. Run multi-hop traversals on every authorisation call to flag suspicious loops.

Knowledge-driven LLM applications

Store documents with embeddings, chunk hierarchy, and citation edges. Plugins call local vector search or external OpenAI APIs to ground answers.

Supply chain visibility

Track parts, suppliers, shipments, and compliance certificates with full lineage. Time-travel queries reconstruct the exact chain for any delivered unit.

Recommendation engines

Traverse user, content, and behavioural signals in real time to deliver personalised recommendations with context.

Real-time that moves the business

Faster time to market

Simplify your architecture and reduce dependencies, allowing teams to ship real-time features faster and iterate with confidence.

Lower TCO

Replace multiple databases, caches, and streaming systems with a single platform - reducing infrastructure, maintenance, and operational costs.

Better user experience

Instant updates and live data synchronisation create responsive, engaging applications that increase user satisfaction and retention.

Future-proof architecture

A flexible, multi-model database with built-in real-time capabilities adapts easily as product requirements and business needs evolve.

TRUSTED BY

Powering innovation across industries

Unlocking insights with knowledge graphs

Samsung uses SurrealDB to power a knowledge graph for real-time audience insights and ad targeting in its ad division.

Learn more

Unified infrastructure monitoring

Tencent uses SurrealDB to consolidate nine backend tools into one real-time monitoring platform.

Learn more

Frequently Asked Questions

How does SurrealDB model graph relationships?

Can I traverse graphs with SurrealQL?

Can I combine graph queries with full-text and vector search?

Does SurrealDB support recursive graph queries?

Do I need a fixed schema for my knowledge graph?

LET'S GOOOO

Ready to build knowledge graphs?

Build dynamic, intelligent systems and enhance AI-driven data analysis, for real-time, actionable insights.