Sparksee
sparsity-technologies.com
- Free plan
- Windows
- macOS
- Linux
- iPhone and iPad
- Android
- API
- Self-hosted
What Sparksee is
Sparksee is an embeddable graph database management system for representing and analyzing relationships among entities. It runs from an application process against a database file, so deployment does not require a cluster or query server. The maker says its compact bitmap-based graph representations support graph computations in mobile and embedded environments. Built-in algorithms include traversal, shortest path, connectivity, community detection, and PageRank. The API supports graph import and export, scripting, and Sparksee Cypher queries, an implementation of openCypher with particularities. Sparksee Algebra is also listed as a query language. API references cover .NET, Java, C++, Objective-C, Python, and openCypher. Sparksee offers disk-level encryption and database recovery after corruption; keeping data on a user's device can support privacy-preserving apps and reduce dependence on cloud connectivity. It supports ACID transactions and Android, iOS, Linux, macOS, and Windows, along with API and self-hosted deployment. Eligible academic users can apply for a free licence. Commercial licence pricing is provided on request and depends on deployment details.
Who it suits
Sparksee suits developers building graph applications for mobile or embedded environments who want an embeddable database rather than a separate query server. Academic users may apply for a free licence with valid school or university identification.
Good to know, and what to watch
- Runs against a database file without a cluster or query server
- Includes traversal, shortest path, connectivity, community detection, and PageRank
- Supports graph import, export, scripting, and Sparksee Cypher
- Offers disk-level encryption and recovery after database corruption
- Supports ACID transactions
- Commercial licence pricing is provided on request
- Academic licence applications require manual validation and valid school or university ID
- Commercial terms depend on deployment details
Plans
Every plan and priceFree trial
Industry
Startup
Academic licence
Free plan · pricing on request
Getting started
- Visit the Sparksee website.
- Choose the academic licence, request a startup or industry licence, or apply for a free trial.
- For an academic licence, submit a valid university or school ID for manual validation.
- Use the developer page's API references for a supported language or query interface.
- Run Sparksee from an application process against a database file on a supported platform.
Limits to check
The free academic licence is restricted to eligible academic applicants and requires ID validation. Commercial licences have no published fixed price; terms are set according to company size and deployment details.
The Club's review
Sparksee is an embeddable graph database for developers building applications around connected data. It is best suited to mobile and embedded projects that need graph analysis close to the data; teams seeking a separately provisioned graph service should look elsewhere.
Overview
Sparksee runs inside an application process and stores its database in a file, avoiding a cluster or query server. That suits software designed to work locally or with limited cloud connectivity, but it also means Sparksee is an embedded component rather than a standalone remote database service.
The maker describes compact bitmap-based graph representations intended to support high-performance computation in mobile and embedded environments. That design is relevant when an application must analyze relationships on a device, though the stated performance focus is not a substitute for evaluating fit with a particular workload.
Key features
Built-in algorithms cover traversal, shortest path, connectivity, community detection, and PageRank. They give developers graph-analysis tools within the database rather than requiring every operation to be built from scratch. Sparksee also supports ACID transactions, graph import and export, scripting, and two query options: Sparksee Cypher (SCL) and Sparksee Algebra.
SCL is described as an implementation of openCypher with particularities. Teams expecting exact openCypher behavior should account for those differences when assessing compatibility. API references cover .NET, Java, C++, Objective-C, Python, and openCypher, giving application developers several language paths to evaluate.
Disk-level encryption is intended to protect data if a device is lost, stolen, or sold, and Sparksee can recover a database after corruption. Local storage can also support privacy-oriented applications and reduce reliance on cloud connectivity. These capabilities are useful for device-based deployments, but they do not by themselves establish an application's broader security or privacy compliance.
Pricing
Sparksee uses a freemium model, with a free plan and commercial licensing on request. The free Academic licence is free for teachers, PhD students, and university research group members. Applicants must provide a valid university or school ID, and approval requires manual validation, so it is not an unrestricted free tier for general commercial development.
The Startup plan has custom pricing for SMEs building apps. The first 10 startups using Sparksee in an app get a 3-year licence waiver, a meaningful opportunity for eligible early adopters, but not a standing free plan for every startup. Industry also has custom pricing, with terms sized to company size and deployment for enterprise-level, real-time graph applications. Commercial licence costs depend on deployment details, including platforms and whether the engine ships to customers, so teams need to request terms for their intended release. A free trial is available; trial applicants receive a build and access to an engineer for modelling questions.
Platforms
Sparksee supports Android, iOS, Linux, macOS, and Windows, as well as API and self-hosted deployments. The combination of mobile and desktop operating systems fits applications that embed graph storage across device types.
Who it's for
Sparksee is a strong fit for developers who need graph data and algorithms inside an application, particularly where local storage, mobile use, or reduced cloud dependence matters. Academic users can apply for a no-cost licence, while startups and enterprises should weigh the deployment-specific commercial terms. It is less suitable for teams whose priority is a ready-to-run graph service or who need commercial pricing known before discussing their deployment.
Pros and cons
- Pro: Embedded, file-based operation avoids provisioning a cluster or query server, which can simplify an application architecture built around local graph data.
- Pro: Built-in traversal, pathfinding, connectivity, community detection, and PageRank provide a substantial set of graph computations within the product.
- Pro: A free academic licence and a three-year waiver for the first 10 qualifying startups create specific routes to lower-cost adoption.
- Con: Commercial terms depend on deployment and require a request, making costs harder to assess upfront.
- Con: SCL has particularities relative to openCypher, which may require adaptation for teams relying on exact language compatibility.
- Con: Its embedded design is not the natural choice for users looking for a separately provisioned graph database service.
Alternatives
For an embedded database focused on local graph workloads, Sparksee is the targeted choice. Consider Memgraph instead if a full in-memory graph database with ACID transactions and on-disk persistence is a better match. NebulaGraph offers an open-source edition under Apache 2.0 with a subset of core features and community support.
TigerGraph is another freemium graph option, with a free trial plan described for a 24-7 service and workspaces up to 32 vCPUs and 256GB RAM. Aerospike Database has a Community Edition described with 8 nodes, 2.5 TB of data, and 2 namespaces. Neo4j offers AuraDB Free as a managed graph database without requiring a payment method, a better fit for readers specifically seeking a managed service.
Quine, TuGraph, and InfiniteGraph are additional graph-database alternatives to compare.
Browse more options in Graph Databases or, for device-oriented storage, Mobile Databases.
Verdict
Choose Sparksee if you are building an application that needs graph storage and analysis embedded on mobile or other local devices: its file-based deployment, algorithms, and language references are its clearest advantages. Look elsewhere if you need a standalone graph service or must settle commercial costs before scoping the deployment.
The verdict
Sparksee provides an embeddable graph database with built-in algorithms, multiple language references, and on-device data storage. Academic users can apply for a free licence; commercial users need to request deployment-specific terms.
Questions people ask
Is Sparksee free?
It offers a free academic licence for eligible teachers, PhD students and university research group members. A valid university or school ID is required, followed by manual validation.
Is there a free trial?
Yes. Trial applicants receive a trial build and access to an engineer for modelling questions.
How much do commercial licences cost?
Commercial licence pricing is provided on request. Terms depend on deployment details, including platforms and whether the engine ships to customers.
Which platforms does Sparksee support?
Its listed platforms are Android, iOS, Linux, macOS and Windows. It is also available through an API and for self-hosted deployment.
What query languages can I use?
Sparksee lists Sparksee Cypher (SCL) and Sparksee Algebra. The API supports Sparksee Cypher, described as an implementation of openCypher with particularities.
What algorithms are included?
The listed algorithms cover traversal, shortest path, connectivity, community detection and PageRank.
What the maker publishes
- Algorithms
- The product page says built-in algorithms are available, and the user manual lists traversal, shortest path, connectivity, community detection, and PageRank algorithms. source
- Algorithms
- Built-in graph algorithms are provided as out-of-the-box tooling for application development. source
- API capabilities
- The API documentation covers database construction, node and edge operations, attributes, collections, import/export, scripting and graph algorithms. source
- Bitmap architecture
- The maker says bitmap-based structures support compact storage and high-performance graph computation in embedded and mobile environments. source
- Company activities
- The company LinkedIn page says Sparsity collaborates on EU-funded projects including LDBC, SoMatch and CoherentPaaS. source
- Company facts
- Crunchbase describes Sparsity Technologies as a private Barcelona spin-off from UPC-Barcelonatech, founded in 2010 by Josep Lluis Larriba Pey, with a listed headcount range of 1–10. source
- Connectivity resilience
- Local graph processing lets mobile analytics continue when cloud connectivity is poor or unavailable. source
- Data and queries
- The API supports graph data import and export, scripting, and Sparksee Cypher queries, described as an implementation of openCypher with particularities. source
- Deployment
- Sparksee runs from an application process against a database file, with no cluster or query server to provision. source
- Deployment classification
- GDB-Engines classifies Sparksee as self-hosted while also supporting embedding in an application. source
- Documentation depth
- Sparsity publishes a starting guide, complete user manual, high-availability material and language-specific reference guides in PDF and browsable HTML. source
- Embedded deployment
- The engine runs inside an application process against a database file, without requiring a cluster or query server. source
- Encryption
- Disk-level encryption is provided to protect data if a device is lost, stolen or sold. source
- Engineering standards
- The developer page states that Sparsity works with ASPICE and MISRA standards. source
- Graph export integration
- The API documentation describes export to Graphviz and GraphML and discusses third-party Blueprints integration. source
- Graph model
- Sparksee models data as a labelled, attributed multigraph with directed or undirected edges, typed objects, attributes and loops. source
- Headquarters
- Barcelona, Spain source
- History
- Wikipedia records development beginning in 2006, a first release in 2008, the DEX name change to Sparksee in February 2014 and version 6.0 in 2021. source
- Independent benchmark limit
- The same study found certain incoming-edge queries on Freebase samples exhausted available RAM and swap space in Sparksee, linked by the authors to a Gremlin implementation problem. source
- Independent benchmark praise
- A microbenchmark study reported Sparksee among the fastest systems for insert, update and delete operations, generally completing them below 100 milliseconds. source
- Independent database profile
- DBDB.io describes Sparksee as a native bitmap-based graph database with full ACID transactions, used especially for social-network analysis. source
- Installation
- Packages are downloaded and unpacked without a general installer, except for an Objective-C MacOS framework installer; Java packages can also be obtained through Apache Maven. source
- Language bindings
- The developer page provides API references for .NET, Java, C++, Objective-C, Python, and openCypher. source
- Language bindings
- Documented bindings are available for .NET, Java, C++, Objective-C and Python, with an openCypher reference. source
Checked against the maker's pages on 15 Oct 2026.