GraphSense is an open-source cryptoasset analytics platform, first developed in 2015 at the Austrian Institute of Technology (AIT) and now maintained in collaboration with the Complexity Science Hub.
Iknaio Cryptoasset Analytics GmbH was founded in 2021 by the creators of GraphSense. It provides GraphSense as a hosted service with near real-time data updates and develops commercial extensions — such as Pathfinder, CaseConnect, QuickLock, and TaxReport — built on top of the open-source platform.
In short: GraphSense is the open-source software stack (MIT-licensed). Iknaio is the company that operates, maintains, and extends it commercially. Development is jointly driven by Iknaio, the Complexity Science Hub, and AIT.
| GraphSense | Iknaio | |
|---|---|---|
| What | Open-source analytics platform | Commercial service & extensions |
| License | MIT | Proprietary (extensions) |
| Access | Self-hosted via GitHub | Hosted at app.iknaio.com / api.iknaio.com |
| Scope | graphsense-lib (analytics engine & REST API), graphsense-dashboard (web UI), TagPacks | Pathfinder, CaseConnect, QuickLock, TaxReport, managed infrastructure |
There are two ways to use GraphSense:
If you do not want to operate your own infrastructure, Iknaio provides GraphSense-as-a-Service with near real-time data:
The complete GraphSense software stack is available on GitHub under the MIT license. Self-hosting gives you full data sovereignty and the ability to run custom analytics jobs.
GraphSense consists of two main components:
graphsense-lib is a Python toolkit that powers the core of the GraphSense platform. It handles blockchain data ingestion, transformation, and serves the REST API.
Requires Python ≥ 3.10 and a Cassandra cluster as storage backend.
Install via uv add graphsense-lib[all].
See the graphsense-lib README
for configuration and deployment details.
graphsense-dashboard is a web-based GUI built in Elm for interactive cryptocurrency analysis. It connects to a GraphSense REST API endpoint and runs entirely in the browser.
Requires NodeJS. Run make serve for development or make build to produce a
production build. Docker Compose deployment is also supported.
See the graphsense-dashboard README
for setup instructions.
GraphSense processes billions of transactions and requires substantial resources. In our production environment we use a Cassandra cluster with:
All nodes are connected via bonded 40 Gbit ethernet. A typical personal computer does not have sufficient resources for a full deployment, but each component can be run individually in development mode.
Iknaio has produced tutorial videos demonstrating how to use the GraphSense dashboard and API.
Check out the tutorials at: https://www.iknaio.com/learning
Q: Can I run GraphSense in the cloud?
Technically yes, but deployment procedures are currently designed for on-premises environments and may require adaptation for cloud providers. We run on-premises because we move large volumes of data (which gets expensive in the cloud) and value full independence from third-party infrastructure.
Q: Does GraphSense support real-time updates?
Yes. Using the ingest and delta-update features of graphsense-lib, a near real-time system can be built. Update frequency depends on the operator — the Iknaio-hosted instance updates data near real-time.
Q: How are entity and abuse types assigned to addresses?
They are assigned manually via GraphSense TagPacks. Labels and tags created by dashboard users are stored on the client side only and are never sent to the server.
Q: Who is funding GraphSense?
Development relies on contributions from the core team at Iknaio, the Complexity Science Hub, and AIT. It has also been supported by public research grants including TITANIUM (EU Horizon 2020), VIRTCRIME (FFG KIRAS), KRYPTOMONITOR (FFG KIRAS), and DeFiTrace (2023–2025).
Q: I need a specific new feature — can you provide it?
File a feature request issue in graphsense-lib or graphsense-dashboard, or write to contact@graphsense.org.