Memgraph
memgraph.com
- Free plan
- Windows
- macOS
- Linux
- Web
- API
- Self-hosted
What Memgraph is
Memgraph is an in-memory graph database and platform for graph computations on streaming data. It supports Cypher and lists Kafka, Pulsar, and Redpanda as streaming connectors. Community Edition includes ACID transactions, on-disk persistence, high-availability replication, and vector search. MAGE provides more than 60 graph algorithms, with integrations for NetworkX, igraph, and cuGraph. Memgraph Lab offers visual tools to query, manage, import, and visualize graph data, plus natural-language querying through GraphChat. The platform lists drivers for languages including Python, C/C++, Rust, Node.js, C#, Go, Java, and Ruby. Enterprise editions add features such as role- and label-based access controls, SSO, authentication options, audit logging, failover, replication, disaster recovery, and monitoring. Community Edition is free forever and includes Discord support. Memgraph Cloud is managed on AWS in six regions, with instance sizes from 1 GB to 32 GB RAM.
Who it suits
Memgraph suits teams building graph applications or working with streaming data, including use cases such as fraud detection, GraphRAG, and network analysis. It offers a free open-source edition as well as enterprise and managed cloud options.
Good to know, and what to watch
- Cypher support with Kafka, Pulsar, and Redpanda connectors
- Community Edition includes ACID transactions and persistence
- MAGE provides more than 60 graph algorithms
- Drivers span a broad set of programming languages
- Cloud instances range from 1 GB to 32 GB RAM
- Community support is through Discord
Plans
Every plan and priceEnterprise Edition
Enterprise Edition — AI Platform
Enterprise Edition — Graph Analytics
OEM / SaaS licence
Free plan · 30-day trial
Getting started
- Choose Community Edition, an Enterprise edition or Memgraph Cloud
- Use the supported deployment platform, including self-hosted systems, web access or AWS-hosted Cloud
- Connect through Cypher and a listed language driver
- Import or visualize data in Memgraph Lab
- Add Kafka, Pulsar or Redpanda for streaming inputs
- Start the free 30-day Enterprise trial if evaluating Enterprise features
Limits to check
Community Edition is free forever. Memgraph Cloud offers 1 GB to 32 GB RAM instances, while Enterprise pricing scales with memory capacity or graph-data capacity depending on the edition.
The Club's review
Memgraph is an in-memory graph database and platform for graph computation on streaming data. It is best suited to teams that need graph queries alongside streaming connectors, algorithms, or vector search. Its strongest case is the capable free self-hosted edition; the main trade-off is custom pricing for enterprise and cloud options.
Overview
Memgraph centers on Cypher-based graph workloads and supports ACID transactions. Community Edition adds on-disk persistence and high-availability replication, so teams can start with a self-hosted database without giving up those core operational capabilities.
The product combines database software with streaming connectors, graph algorithms, vector search, and visual tooling. Memgraph names AI infrastructure, GraphRAG, AI memory, fraud detection, network analysis, supply-chain dependencies, and infrastructure monitoring among its enterprise use cases. Founded in 2016, the company is headquartered in Sevenoaks, Kent, United Kingdom. Buyers comparing Graph Databases should consider whether its streaming focus and operational features match their workload.
Key features
Streaming and algorithms
Kafka, Pulsar, and Redpanda connectors link Memgraph to streaming systems. MAGE offers 60+ graph algorithms and integrations with NetworkX, igraph, and cuGraph. That combination is suited to teams that need graph computation close to changing data, rather than a graph store used only for occasional queries.
Querying and developer access
Cypher is the query language, and Memgraph lists programmatic drivers for Python, C/C++, Rust, Node.js, C#, Go, Haskell, Java, JavaScript, PHP, and Ruby. The range makes it practical to connect from varied application stacks without relying on one language ecosystem.
Visual tools and vector search
Memgraph Lab provides a visual interface for querying, managing, importing, and visualizing graph data. Its GraphChat feature supports natural-language querying, which can make exploration more approachable for users who do not want to begin with handwritten queries. Vector search is also included among the product capabilities.
Enterprise operations
Enterprise editions add role- and label-based access controls, SSO, LDAP/PAM authentication, TLS encryption in transit, and query audit logging. Availability features include automatic failover, synchronous and asynchronous replication, disaster recovery, and no-downtime updates. Prometheus metrics and query and transaction latency tracking support operational monitoring; dedicated engineering help is provided through a Slack channel. These controls matter for production teams, but they sit behind custom-priced plans rather than the free edition.
Pricing
Memgraph uses a freemium model, with a free plan and a 30-day Enterprise trial. The trial is a useful way to evaluate enterprise capabilities, but it does not establish the ongoing cost of a paid deployment.
| Plan | Price | What it includes |
|---|---|---|
| Community Edition | Free | Free forever, open source, full in-memory graph database, ACID transactions with on-disk persistence, high-availability replication, Cypher, streaming connectors, vector search, MAGE algorithms, and Discord community support. |
| Enterprise Edition — Graph Analytics | Custom pricing | Price scales with total memory capacity, counting graph data and vector indexes together. Includes enterprise security and access controls, SSO, multi-tenancy, automatic failover, disaster recovery, audit logging, Prometheus monitoring, and dedicated engineering support. |
| Enterprise Edition — AI Platform | Custom pricing | Price scales with graph data capacity only. Includes unlimited vector indexes, expanding AI tooling, enterprise security and access controls, multi-tenancy, automatic failover, disaster recovery, audit logging, Prometheus monitoring, and dedicated engineering support. |
| Memgraph Cloud | Custom pricing | Fully managed AWS hosting in six regions, instances from 1 GB to 32 GB RAM, automatic updates and backups, and community support. |
Community Edition is the clear fit for teams able to run the database themselves and comfortable with Discord support. Enterprise is aimed at organizations that need security, multi-tenancy, continuity, monitoring, or direct engineering support; the Graph Analytics plan counts vector indexes toward its memory-based pricing, while the AI Platform scales on graph data capacity and includes unlimited vector indexes. Cloud shifts operation to a managed service and includes updates and backups, but caps instance choices at 32 GB RAM and provides community rather than dedicated engineering support.
Platforms
Memgraph supports API access, Linux, macOS, Windows, self-hosted deployment, and web access. The self-hosted option suits teams that want control over deployment; Memgraph Cloud is the managed alternative for teams preferring AWS hosting.
Who it's for
Memgraph suits developers and data teams working with streaming graph workloads, or organizations that need graph algorithms and vector search in the same platform. The free edition is substantial enough for teams to begin with transactions, persistence, replication, and connectors. Teams with strict access-control, availability, or support requirements should evaluate Enterprise; those seeking managed operations can consider Cloud. It is less compelling for buyers who do not need graph computation or streaming integrations and simply want a conventional database.
Pros and cons
- Pros: Community Edition combines ACID transactions, on-disk persistence, high-availability replication, streaming connectors, vector search, and algorithms at no cost.
- Pros: Named connectors and broad language drivers make it adaptable to streaming pipelines and varied application stacks.
- Pros: Lab brings visual query, import, management, and visualization tools together, with GraphChat for natural-language querying.
- Cons: Enterprise and Cloud have custom pricing, so buyers cannot compare ongoing cost from a fixed published price.
- Cons: Community support is Discord-based; dedicated engineering support is reserved for Enterprise.
- Cons: Cloud instances range only from 1 GB to 32 GB RAM, which may not fit workloads needing larger managed instances.
Alternatives
Sparksee is another freemium option with an academic licence offered free to qualifying users who can provide a valid school or university ID; it also supports mobile platforms, unlike Memgraph's listed platforms.
NebulaGraph is worth considering if an Apache 2.0 open-source edition and community collaboration are priorities; its free edition is described as a subset of core features.
TigerGraph offers a free Savanna trial with a stated $0 per GB/month* rate for 24-7 service in US/Tier 1 and workspace limits of up to 32 vCPUs and 256GB RAM; consider it if those cloud terms suit the workload.
Aerospike Database may suit teams comparing a database with a Community Edition stated to support 8 nodes, 2.5 TB of data, and 2 namespaces.
Neo4j is an alternative with a free managed AuraDB plan that requires no credit card or other payment method.
Quine is an open-source option from thatDot for buyers considering a graph database project rather than Memgraph's broader mix of plans and tooling.
TuGraph offers a free Community Edition described as centralized, with high-availability deployment and TB-scale data; it is worth comparing where those characteristics are a priority.
InfiniteGraph provides a free version with up to 50GB of disk storage, a useful point of comparison for teams assessing storage needs.
Verdict
Choose Memgraph if your team needs graph computation over streaming data and wants Cypher, algorithms, vector search, and a usable free self-hosted starting point. Its most persuasive advantage is that the Community Edition includes persistence and high-availability replication alongside core query and streaming features. Look elsewhere if you need a published paid price, larger managed-cloud instances, or a graph database without Memgraph's streaming-oriented scope.
The verdict
Memgraph combines streaming graph computation with query, visualization, and algorithm tools. Community Edition offers a free starting point; enterprise capabilities and the managed cloud option address different deployment needs.
Questions people ask
Is Memgraph open source?
Yes. Community Edition is open source and free forever.
What query language does it use?
Memgraph supports the Cypher query language.
Does it offer a free trial?
Yes. Memgraph offers a free 30-day Enterprise trial.
What platforms are available?
Platforms listed include API, Linux, macOS, self-hosted, web and Windows. Memgraph Cloud is AWS-hosted in six regions.
What support is included?
Community Edition lists Discord support. Enterprise includes direct engineering support through a dedicated Slack channel.
What the maker publishes
- Agentic AI
- Reasoning graphs expose agent action paths and make traversed evidence inspectable. source
- AI integrations
- The documentation provides LangChain and LangGraph tooling, including a Memgraph toolkit and MemgraphQAChain. source
- AI memory
- Memgraph models semantic, episodic and procedural long-term memory as a queryable graph. source
- Algorithms
- MAGE provides 60+ graph algorithms, with integrations for NetworkX, igraph, and cuGraph. source
- Availability
- Enterprise includes automatic failover, synchronous and asynchronous replication, disaster recovery, and no-downtime updates. source
- Built-in algorithms
- Core traversal procedures include breadth-first search, depth-first search, weighted shortest path and all shortest paths. source
- Client libraries
- Official documentation lists clients or guides for C#, C/C++, Go, GraphQL, Haskell, Java, JavaScript, Node.js, PHP, Python, Ruby, Rust and Swift. source
- Cloud infrastructure
- Memgraph Cloud is fully managed on AWS, offers six regions, and supports instances up to 32 GB RAM and eight CPU cores. source
- Cloud limits
- Memgraph Cloud lists instance sizes from 1 GB to 32 GB RAM and says the service is AWS hosted in six regions. source
- Cloud security
- Cloud connections use SSL encryption and Cloud instances include Enterprise role-based access control. source
- Company headquarters and founding
- LinkedIn lists Memgraph as headquartered in London, England, founded in 2016, privately held and employing 11–50 people. source
- Company history
- Memgraph’s about page says its founders celebrated the company’s sixth year in 2022. source
- Company legal name
- Memgraph’s product pages identify the company as Memgraph Ltd. source
- Custom modules
- Developers can extend Memgraph with query modules written in C, C++, Python or Rust, with optional NVIDIA cuGraph acceleration. source
- Deployment
- Memgraph supports Docker, native Linux packages, Kubernetes and cloud deployments on AWS, Azure and GCP. source
- Documentation depth
- The docs cover installation, client libraries, migration from graph and SQL systems, query modules, deployment, configuration, troubleshooting and release notes, and expose a public MCP documentation server. source
- Drivers
- The capabilities page lists programmatic drivers for Python, C/C++, Rust, Node.js, C#, Go, Haskell, Java, JavaScript, PHP, and Ruby. source
- Founded
- 2016 source
- Funding
- Memgraph announced a $9.34 million round led by Microsoft’s M12 venture fund with Heavybit, In-Q-Tel, Counterview Capital, ID4 Ventures and Mundi Ventures participating. source
- GraphRAG
- The product combines vector or text search with multi-hop graph traversal so retrieval can follow connected entities rather than similarity alone. source
- Headquarters
- Sevenoaks, Kent, United Kingdom source
- Independent licence assessment
- Is It Really FOSS concludes that Memgraph is not FOSS because BSL 1.1 and the Memgraph Enterprise License do not provide FOSS rights, despite open-source marketing. source
- Independent review limitation
- Hysen Labs notes that benchmark speed claims are vendor statements until reproduced and flags the mixed licence model as an adoption cost. source
- Installation platforms
- The download page provides one-command installers for Linux/macOS and PowerShell installation for Windows. source
Checked against the maker's pages on 15 Oct 2026.