Today, we published our blog post What’s Next: A Look at Unity’s 2025 Roadmap and our roadmap presentation from earlier this week at GDC, where we shared an overview of the Unity Engine roadmap for 2025.
We highlighted our commitment to making Unity more stable and production-tested for game development and live operation for all users. We also provided clarity on how Unity 6.0 will be supported, a preview of what’s coming in Unity 6.1, and a look ahead to what’s next.
Additionally, our team has created a number of dedicated threads here on Unity Discussions to help facilitate discussions about what we shared today:
Lots of focus on services - AI services, multiplayer services, analytics, etc. Very few updates on actual functionality, which is what I’d except from a roadmap. The video also covers mostly features available right now or releasing in a few months with 6.1 so it’s more a recap than a roadmap.
I’m still wondering where all the Unity 7 hype disappeared. It was a big step in the right direction, but then just puffed out.
I understand that so much of this stuff is being pushed by higher ups (android because of the mobile market, VR because it’s the “new kind of gaming”, etc.), but I really, really don’t care about those improvements, and I know of at least nine other Unity devs both in the market and working on their own that don’t as well.
I even have a buddy that makes mobile apps with Unity that wishes you guys would stop focusing on that “niche” and make the engine as a whole better and/or more modern.
The graphical optimizations are great. I realize those are probably hard to slip in since they’re “boring” and don’t catch eyes, so I appreciate it.
What I’m really waiting for, to be totally honest, is CoreCLR. I have been told this is two years away for ten years. It’s insane that it’s taken this long or, for that matter, that it isn’t the immediate and most prominent priority of the whole engine at this point.
In short, I was disappointed, but I wasn’t surprised. I’ll be excited when CoreCLR is in an alpha, or when the rendering pipelines are actually unified.
It was kinda disapointing. I feel the development is slowed down.
I’m suprised to see the new animation system is coming in 6.x version.
Also, Unified Rendering wasn’t mentioned. Is it still happening, right? RIGHT?
Lastly, .NET Modernization was mentioned too little like it’s too far away. Each year it will become harder to do it. We need more transparency about this.
I can speak for only one portion of this work (Content Pipeline) but the longer-term foundational/modernization work is still in progress and on our longer-term roadmap.
We want to get this new tech in users’ hands as quickly as we can because we’re very excited about the benefits it will deliver in terms of iteration time, performance, and reduction of overlapping/redundant systems. But as you can imagine with engine areas as foundational as asset importing, loading, building, and downloading/distribution, modernizing these areas of the engine is quite an undertaking. More important than delivering this tech quickly is delivering this work responsibly, validating performance improvements & stability with developers along the way.
So we will have much more detail to share later, but know we are still very focused on things like:
Reducing the amount of time users are blocked by things like asset imports and content builds
Eliminating the confusion of several overlapping systems and pipelines with similar purposes - and increasing our own speed in delivering continual improvements by focusing on fewer systems
Reducing the burden of inspecting & optimizing complex dependency trees in your project’s built content (e.g. in Assetbundles/Addressables)
Giving more fine-grained, reliable control over asset unloading at runtime
Embracing a more multi-threaded, multi-process architecture to allow for scalable performance at edit time & runtime
Reducing unnecessarily large content downloads for your players by embracing more granular/incremental content build artifacts
I am looking forward quite a lot to the included asset generators for sound and even 3d models. I know this is controversial but also technically very exciting and I think especially for prototyping this can be super valuable tooling.
What I would urge Unity to do is to provide a native MCP server exposing the Unity API and then allow all kinds of third parties to go crazy with integrations. This could open up a lot of AI craziness, the good, the bad and the amazing ones.
I wish official can prioritize merging Addressable and ECS content management together since currently there is no proper tooling like Addressable for ECS land until today. And also for [IN-61082] Size of the Content Update generated folder increases when changes are made to the Scene, official really to fix soon than later since it’s last year September 2024 bug report.
