DBMS > Milvus vs. Neo4j vs. Pinecone vs. ScyllaDB vs. Snowflake

System Properties Comparison Milvus vs. Neo4j vs. Pinecone vs. ScyllaDB vs. Snowflake

Editorial information provided by DB-Engines

Name

Description

A DBMS designed for efficient storage of vector data and vector similarity searches

Scalable, ACID-compliant graph database designed with a high-performance distributed cluster architecture, available in self-hosted and cloud offerings

A managed, cloud-native vector database

Cassandra and DynamoDB compatible wide column store

Cloud-based data warehousing service for structured and semi-structured data

Primary database model

Secondary database models

Key-value store

DB-Engines Ranking measures the popularity of database management systems
Trend Chart

Score	2.78
Rank	#103	Overall
	#3	Vector DBMS

Score	44.89
Rank	#21	Overall
	#1	Graph DBMS

Score	3.23
Rank	#92	Overall
	#2	Vector DBMS

Score	4.08
Rank	#76	Overall
	#5	Wide column stores

Score	130.36
Rank	#8	Overall
	#5	Relational DBMS

Website

Technical documentation

milvus.io/docs/overview.md

neo4j.com/docs

docs.pinecone.io/docs/overview

docs.scylladb.com

docs.snowflake.net/manuals/index.html

Developer

Neo4j, Inc.

Pinecone Systems, Inc

ScyllaDB

Snowflake Computing Inc.

Initial release

2019

2007

2019

2015

2014

Current release

2.3.4, January 2024

5.20, May 2024

ScyllaDB Open Source 5.4.1, January 2024

License

Commercial or Open Source

Open Source

Apache Version 2.0

Open Source

GPL version3, commercial licenses available

commercial

Open Source

Open Source (AGPL), commercial license available

commercial

Cloud-based only

Only available as a cloud service

yes

DBaaS offerings (sponsored links)

Database as a Service

Providers of DBaaS offerings, please contact us to be listed.

Zilliz Cloud – Cloud-native service for Milvus

Neo4j Aura: Neo4j’s fully managed cloud service: The zero-admin, always-on graph database for cloud developers.

Scylla Cloud: Create real-time applications that run at global scale with Scylla Cloud, the industry’s most powerful NoSQL DBaaS

Implementation language

C++, Go

Java, Scala

C++

Server operating systems

Linux
macOS

10.14 or later
Windows

with WSL 2 enabled

Linux

Can also be used server-less as embedded Java database.
OS X
Solaris
Windows

hosted

Linux

hosted

Data scheme

schema-free and schema-optional

schema-free

yes

support of semi-structured data formats (JSON, XML, Avro)

Typing

predefined data types such as float or date

Vector, Numeric and String

yes

String, Number, Boolean

yes

XML support

Some form of processing data in XML format, e.g. support for XML data structures, and/or support for XPath, XQuery or XSLT.

yes

Secondary indexes

yes

pluggable indexing subsystem, by default Apache Lucene

yes

cluster global secondary indices

SQL

Support of SQL

SQL-like DML and DDL statements (CQL)

yes

APIs and other access methods

RESTful HTTP API

Bolt protocol
Cypher query language
Java API
Neo4j-OGM

Object Graph Mapper
RESTful HTTP API
Spring Data Neo4j
TinkerPop 3

RESTful HTTP API

Proprietary protocol (CQL)

compatible with CQL (Cassandra Query Language, an SQL-like language)
RESTful HTTP API (DynamoDB compatible)
Thrift

CLI Client
JDBC
ODBC

Supported programming languages

C++
Go
Java
JavaScript (Node.js)
Python

.Net
Clojure
Elixir
Go
Groovy
Haskell
Java
JavaScript
Perl
PHP
Python
Ruby
Scala

Python

For CQL interface: C#, C++, Clojure, Erlang, Go, Haskell, Java, JavaScript, Node.js, Perl, PHP, Python, Ruby, Rust, Scala
For DynamoDB interface: .Net, ColdFusion, Erlang, Groovy, Java, JavaScript, Perl, PHP, Python, Ruby

JavaScript (Node.js)
Python

Server-side scripts

Stored procedures

yes

User defined Procedures and Functions

yes, Lua

user defined functions

Triggers

yes

via event handler

similar concept for controling cloud resources

Partitioning methods

Methods for storing different data on different nodes

Sharding

yes using Neo4j Fabric

Sharding

yes

Replication methods

Methods for redundantly storing data on multiple nodes

Causal Clustering using Raft protocol

available in in Enterprise Version only

selectable replication factor

Representation of geographical distribution of servers is possible

yes

MapReduce

Offers an API for user-defined Map/Reduce methods

Consistency concepts

Methods to ensure consistency in a distributed system

Bounded Staleness
Eventual Consistency
Immediate Consistency
Session Consistency
Tunable Consistency

Causal and Eventual Consistency configurable in Causal Cluster setup
Immediate Consistency in stand-alone mode

Eventual Consistency
Tunable Consistency

can be individually decided for each write operation

Immediate Consistency

Foreign keys

Referential integrity

yes

Relationships in graphs

yes

Transaction concepts

Support to ensure data integrity after non-atomic manipulations of data

ACID

Atomicity and isolation are supported for single operations

ACID

Concurrency

Support for concurrent manipulation of data

yes

Durability

Support for making data persistent

yes

In-memory capabilities

Is there an option to define some or all structures to be held in-memory only.

yes

in-memory tables

User concepts

Access control

Role based access control and fine grained access rights

Users, roles and permissions. Pluggable authentication with supported standards (LDAP, Active Directory, Kerberos)

Access rights for users can be defined per object

Users with fine-grained authorization concept, user roles and pluggable authentication

More information provided by the system vendor

Specific characteristics

Milvus is an open-source and cloud-native vector database built for production-ready...
» more

Neo4j delivers graph technology that has been battle tested for performance and scale...
» more

ScyllaDB is engineered to deliver predictable performance at scale. It’s adopted...
» more

Competitive advantages

Highly available, versatile, and robust with millisecond latency. Supports batch...
» more

Neo4j is the market leader, graph database category creator, and the most widely...
» more

Highly-performant (efficiently utilizes full resources of a node and network; millions...
» more

Typical application scenarios

RAG: retrieval augmented generation Video media : video understanding, video deduplication....
» more

Real-Time Recommendations Master Data Management Identity and Access Management Network...
» more

ScyllaDB is ideal for applications that require high throughput and low latency at...
» more

Key customers

Milvus is trusted by thousands of enterprises, including PayPal, eBay, IKEA, LINE,...
» more

Over 800 commercial customers and over 4300 startups use Neo4j. Flagship customers...
» more

Discord, Epic Games, Expedia, Zillow, Comcast, Disney+ Hotstar, Samsung, ShareChat,...
» more

Market metrics

As of January 2024, 25k+ GitHub stars 10M+ downloads and installations 3k+ enterprise...
» more

Neo4j boasts the world's largest graph database ecosystem with more than 140 million...
» more

ScyllaDB typically offers ~75% total cost of ownership savings, with ~5X higher throughput...
» more

Licensing and pricing models

Milvus was released under the open-source Apache License 2.0 in October 2019. Fully-managed...
» more

GPL v3 license that can be used all the places where you might use MySQL. Neo4j Commercial...
» more

ScyllaDB Open Source - free open source software (AGPL) ScyllaDB Enterprise - subscription-based...
» more

News

This Week in Neo4j: Podcast, Testing, Knowledge Graph, GenAI and more
8 June 2024

Neo4j and Snowflake Bring Graph Data Science Into the AI Data Cloud
4 June 2024

RDF vs. Property Graphs: Choosing the Right Approach for Implementing a Knowledge Graph
4 June 2024

This Week in Neo4j: Importing Data, NODES, GenAI, Going Meta and more
1 June 2024

openCypher Will Pave the Road to GQL for Cypher Implementers
22 May 2024

We invite representatives of system vendors to contact us for updating and extending the system information,
and for displaying vendor-provided information such as key customers, competitive advantages and market metrics.