In the rapidly evolving landscape of data management, Neo4j is once again setting the standard for innovation and user-centric solutions. As organizations increasingly rely on graph databases to power their most critical applications, the ability to adapt and scale without disruption is paramount. Today, we’re excited to announce a series of transformative changes that will not only enhance the flexibility and stability of our platform but also solidify Neo4j’s position as the industry-leading graph database solution.
Listening to Our Customers: Addressing Core Challenges
We’ve heard from our users — whether database administrators, security professionals, or application developers — that upgrading Neo4j databases can be a complex and sometimes disruptive process across multiple teams. Concerns range from the challenge of migrating applications to the need for immediate access to the latest security fixes and features. Our goal is to make these processes seamless, empowering teams to innovate without fear of disruption.
Decoupling Database and Language Versions: A Game-Changer
One significant change we’re introducing is the decoupling of server versions from the Cypher query language versions. Today we’re announcing Cypher Language Version 25. Here’s the game-changer: Users will be able to choose to use either the current Cypher 5 or Cypher 25, independent of the server version they’re running, and this can even be decided per query!
This decoupling allows for unparalleled flexibility:
- Database administrators can upgrade their databases without forcing immediate changes to applications.
- Application developers can experiment with new Cypher features as soon as they’re available, without the need for a complete application overhaul.
- Security administrators can apply the latest security patches to Neo4j, even those that are major changes to the platform, without waiting for application teams to adapt.
Cypher 25 will be released as a new capability after the Long-Term Support (LTS) of version 5 of the Neo4j database, planned to be made available later this year in November. The Cypher API will be semantically versioned, with version 25 being a nod to its 2025 release date. With this approach, Neo4j is empowering teams to move at their own pace while ensuring they have access to the latest and greatest in graph database technology.
The Cypher API Compatibility Layer: Seamless Transitions
One of the standout features of our new approach is the Cypher API compatibility layer. This allows organizations to specify the default Cypher version across the database or per query, making it easier to transition applications to newer versions of the query language without disruption. New capabilities will be introduced in the latest Cypher 25, while Cypher 5 remains frozen. Whether you’re ready to embrace Cypher 25 or prefer the stability of Cypher 5, the choice is yours.
Predictable and Flexible Server Updates: Introducing Calendar Versioning
Since the first release of Neo4j, every few years we’ve released a new major version like 4.0 and 5.0. Pivotal new features and performance improvements were introduced only in these major version releases, but they also came with a complex migration process that would require downtime. We don’t believe this is beneficial for our customers and have taken a different approach. For instance, rather than waiting for a 6.0 release, we introduced a significant improvement with Block Storage in version 5.14. This approach, while beneficial in getting new capabilities to our users faster, has occasionally raised questions about how we’ll handle server versioning in the future.
To further streamline our release process we’re making some changes along with a commitment to enabling online rolling upgrades to future versions. These changes better align to how we release our product features and how customers update their database. To mark this transition away from major versions, we’re adopting a calendar-based versioning scheme. Gone are the days of traditional semantic versioning. After the release of version 5 LTS, new Neo4j database releases will be versioned with the format YYYY.MM.Patch in a continuous release chain. For example, a server release in January 2025 will be versioned as 2025.01.00.
This change brings several benefits:
- Regular access to new features – Monthly production-ready releases packed with innovations will keep the Neo4j environment on the cutting edge, with the latest features and improvements delivered consistently and unconstrained by version numbers.
- Simplified upgrade path – Our new versioning scheme makes it easier to predict, plan, and carry out upgrades, reducing the complexity for administrators.
- Optional long-term support – Every release is production-ready and supported via the simplified upgrade path. For self-managed customers who are unable to update, we offer our Long-Term Support (LTS) versions. These are targeted to be released every 18-24 months and carry several years of support. The goal will be to allow direct upgrades between a LTS version and the next LTS version. LTS versions will be released when we are comfortable with the feature set and the time is right.
The Neo4j client drivers benefit from a different release process to the server so we believe that keeping them on their current semantic version numbers will be the best option for customers. We will provide a documented support matrix making it clear which versions of the server are compatible with which driver.
A Thoughtful Approach to Deprecation: Ensuring a Smooth Transition
At Neo4j, we understand that change can be challenging, especially when it involves features that have become integral to your workflows. That’s why we have a deprecation policy that prioritizes customer’s needs and minimizes disruption. Removing features and implementing breaking changes in a database platform, while initially disruptive, can lead to significant long-term benefits. It allows the platform to evolve, shedding outdated or inefficient functionalities that may hinder performance, security, or usability. Our approach is clear and deliberate: We try to avoid deprecating features, but when needed, a feature will be deprecated only when an alternative solution is fully in place.
When a feature is deprecated it remains fully supported, but we are sharing our intention to remove the capability in a future release, and users should plan to use an alternative solution. With the new approach of server versioning, we have some straightforward guidance; we will provide at least 12 months’ notice before any feature is removed from future releases, ensuring that users have ample time to adapt. Additionally, database features will not be removed until after an LTS version of the database has been released.
These processes will be put in place from 2025 onward, and with our proactive notifications, users are informed every step of the way. This careful, customer-focused deprecation process ensures that updating to newer versions of Neo4j is as seamless and predictable as possible, allowing customers to plan and upgrade with confidence.
Looking Ahead: A Future-Ready Database
Neo4j’s commitment to innovation doesn’t stop here. By decoupling the Cypher query language from the core database and introducing calendar-based versioning, we’re laying the groundwork for even more dynamic, responsive, and future-proof database solutions. This new approach allows us to release large compelling features as they become ready, rather than waiting for a major database version update, ensuring that Neo4j remains at the forefront of graph database technology.
Summary
With these exciting changes, Neo4j is once again leading the way in the graph database space. By addressing the key challenges faced by our customers, we’re not just making upgrades easier but we’re redefining what’s possible with a graph database. Whether you’re a database administrator, security professional, or application developer, Neo4j is here to empower you with the tools and flexibility you need to innovate without limits.
Stay tuned as we continue to push the boundaries of what’s possible with graph databases. The future of data management is here, and it’s powered by Neo4j.
