UI Toolkit development status and next milestones – November 2025

Hi everyone,

Coming out of Unite and the Unity Engine roadmap session, we’d like to share a UI Toolkit progress update: what’s new in 6.3, what’s actively in development, and where we need your input on what to tackle next.

:sparkles: Summary of what’s new in Unity 6.3

Unity 6.3 brings a packed set of upgrades to UI Toolkit, spanning rendering, authoring, testing, and text, so you can build sharper, faster, and more maintainable interfaces.

  • Vector graphics for UI: Native SVG import and rendering in UI Toolkit. Use scalable assets that stay sharp at any resolution, with Painter2D for antialiased results. Configure assets as UI Toolkit Vector Images and apply them in UI Builder.
  • Custom shaders for UI Toolkit via Shader Graph: New UI target and Render Type Branch node to control rendering per mesh type. Materials can be assigned in code, USS, or UI Builder. Available in 6.3 for URP.
  • Post-processing filters via USS: Built-in filters (blur, grayscale, invert, opacity, sepia, tint, hue-rotate, contrast) plus custom filter definitions. Filters apply to subtrees for localized effects.
  • World-space UI (in 6.2): Runtime UI in 3D space with proper sizing, transforms, and interaction, expanded by 6.3’s rendering and styling updates.
  • Quality-of-life in authoring: Image control in UI Builder, aspect-ratio property for layouts, auto-resize text, and extensibility APIs for custom controls.
  • UI testing framework: UI Test Fixtures for controlling the update loop, simulating interactions, and validating hierarchy or visual state locally and in CI.
  • Advanced Text Generator (ATG) migration continued with UI Toolkit in the Editor itself now using ATG by default with full unicode and bi-directional (inc. RTL) text support.

:construction: What’s in active development

Here’s what we’re actively building next. These features aim to make UI Toolkit faster to author, better for animation, broader in styling via standard USS, and easier to ship at scale.

  • Advanced animation workflows: Author keyframed animations for VisualElements directly in the Animation window, with in-context previews.
  • Scene-view UI authoring: Place, select, and manipulate UI in the Scene view using the new Scene Hierarchy. Supports animation workflows and world-space UI.
  • Referencing of Visual Element defined in UXML, from MonoBehaviour scripts.
  • Additional improvements to UI Toolkit custom shaders, such as support for vertex data
  • Visual polish with standard USS properties: Gradients, drop shadows, and blur or backdrop-style effects.
  • Performance and profiling focus: Work to keep UI Toolkit fast and predictable, plus improved tooling to help you diagnose bottlenecks.
  • Continued development of Advanced Text Generator (ATG) parity, moving IMGUI and UI Toolkit runtime defaults to ATG, bringing full unicode and bi-directional (inc. RTL) text support to both Editor and Runtime.

We aim to land the features above by Unity 6.7 LTS, with increments arriving in supported updates between now and then.

:light_bulb: What should we prioritize next?

We’re planning beyond Unity 6.7 and want your guidance. The most-requested areas we’ve heard lately include:

  • Figma-to-Unity import for UI Toolkit
  • Particles and 3D objects in UI
  • Soft masking
  • Addressable Asset Bundles support
  • Grid layout aligned with CSS Grid
  • Additional CSS properties and selectors
    • Z-index, gap, media queries; pseudo-classes such as :first-child, :last-child, :nth-child
    • Additional units, such as vh/vw and em/rem
  • Data bindings improvements
    • Event binding in UXML
    • Add support for non-attribute bindings, such as itemsSource, and UXMLObjects
    • Mock data source workflows
    • Streamline ListView bindings
    • Bindings debugging workflows
  • APIs to manipulate UI assets and load UXML/USS at runtime

Tell us what matters most for your production workflows. Which items unblock you fastest? If we missed related needs, please share specifics.

:speech_balloon: Share feedback

We’d love to hear what you need next and what you’re building with the latest features. If you’ve adopted vector graphics, filters, custom shaders, or world-space UI, tell us how they’re working for you. If you’re planning a migration or prototyping new interfaces, share what you hope to build and which features would help you get there fastest. Examples, priorities, and constraints help us focus on what matters most.

Thanks for building with UI Toolkit. Your feedback directly shapes what lands in supported updates and next year’s LTS.

*Edit (Dec 2, 2025): Updated the UI Test Framework link to the dedicated post — UI Test Framework :up_right_arrow:

29 Likes

For feedback on what next:

The data bindings improvements would be the most welcomed.
It would allow devs to do a lot more. Specially with the event Binding in UXMLand the non-attribute bindings. Having a more streamlined ListView bindings set up would be a good qol list for people creating data editors.

Example data editor for RPG characters, inventory items, and room data.

The API to manipulate UI assets and load UXML/USS at runtime would honestly be a really amazing QOL for anyone using UI Toolkit in game scenes.

Honestly the CSS properties being more closer to streamline web CSS while having Unity’s own tweaks would be really good to. Z-Index would be useful for certain editor tools.

1 Like

Data bindings improvements

This is great news. After just going through this subject setting up some in-game list views, the workflow was more cumbersome and difficult than it needed to be. The ability to set mock data for preview in the UI builder that does not affect/override runtime data will be a helpful feature.

I’m not trying to derail this conversation, and I’m absolutely not discounting your feedback; but I thought this was a good opportunity to share the fact that the kinds of Editor tools you are mentioning are trivial to implement using Search. This is not a well known workflow, but it’s super convenient and can quickly replace most custom implemented data editors. One day, this documentation will be easier to find, I promise :smile:

Cheers!

7 Likes

Those planned features all look good to me. If I have to prioritize the top 3 from that list, it would probably be:

  • Grid layout aligned with CSS grid - Currently have to use App UI for this feature. Not bad in of itself but I think it should be something that should be in base UITK. Useful for inventory systems and mobile UI
  • Additional CSS properties and selectors - Those pseudo classes are really useful for styling lists of any kind. Currently have to workaround it through script which isn’t ideal. The additional units are really useful on different mobile resolutions too
  • Data bindings improvements - Others already mentioned it but this would greatly ease the churn for binding, particularly the ListView (and friends) bindings

This is bloody fantastic! The page is updated back in 2023, wish I knew it earlier.

Whoa, game changer.

Will visual elements also gets to reference other visual elements?

1 Like

Soft masking is something pretty important for me

2 Likes

Definitely! Does this include 2D GO?

I have prefabs a player can wear in level and instead of making them twice I show them in the equip menu. It takes a lot of custom code to implement them and I have to potentially discard features like a background behind them since they render behind UI.

Any way to add the ability to select UI elements from game view ? (like ctrl + right click there is now in the scene view), as an opt-in as some people will need to have this shortcut for their game. And both yes to world space UI and performant UI animation.

I suppose it depends on which direction you’re coming at updates from (UGUI crowd vs UI/UX/Web crowd), but to me I think UI Builder could add a lot of value and help users transition from legacy systems if it were more stable and properly supported all element classes. Currently it’s often not wysiwyg and does not really allow designers to easily build content in it dependably. I regularly find that components don’t appear as expected in UIBuilder but look fine in the environment and vice versa.

If UI Builder worked better it would greatly improve adoption from the UGUI user base. I think this is a very important bridge that needs to be prioritized sooner than later and the current UI Builder is not there yet. Frankly, UI Elements is difficult. The success and praise that UGUI got was because users found it to be easy, powerful, and convenient. Those are rarely words I would use to describe UI Elements, but if UI Builder could bridge that gap better then the tone would certainly change for the better.

5 Likes

Grid layout aligned with CSS Grid would be number one on wishlist. If you are already using UI Toolkit you probably have already worked around it, but it would benefit newcomers for sure.

I’d personally like to see more improvements to the navigation and focus systems.
For navigation we really need something closer to UGUI’s navigation, where you can manually define transitions between UI elements.
For focus I want better API: things like disabling Tab focus navigation, preventing an element from losing focus when clicking outside it or constraining focus within a specific window should be simple one-liners instead of the current hell of event registration, trying to understand event bubbling and propagation.

6 Likes

There are features we can work around without being advanced users, and then there are features where workarounds are much harder. Here’s a priority list based on what is a blocker for me and for others who are in the same situation:

  1. Additional improvements to UI Toolkit custom shaders, such as support for vertex data

  2. Visual polish with standard USS properties: Gradients, drop shadows, and blur or backdrop-style effects.

  3. GLOW filter

  4. Advanced Text Generator (ATG)

  5. Particles (VFX Graph) and 3D objects in UI

  6. APIs to manipulate UI assets and load UXML/USS at runtime

  7. Soft masking

  8. Grid layout aligned with CSS Grid

  9. Additional CSS properties and selectors

  10. Z-index, gap, media queries; pseudo-classes such as :first-child, :last-child, :nth-child

1 Like

The in-scene editor will probably help a lot with this. You’ll have set up the exact size and the exact settings for the UI, in the scene view already, so there’s no need to try to match it in the builder.

1 Like

I’m a big proponent of this! Our designers really like to build using figma, so if we could have a semi-automatic import from figma instead of having to hand-match the pixels, we’d save a lot of time.

2 Likes

For me, the main priorities are (in no particular order):

  • Add more missing CSS features like grid, gap, pseudo-classes. These are great QoL improvements and make working with UITk much more enjoyable.

  • Add support for particle effects over the UI. This is common for game UIs, to make the UI feel more responsive and to add crucial feedback for players (e.g. coins earned in the game flying to the coins counter with some particles).

  • One great thing about IMGUI is that it allows creating small single-file tools that can be easily dropped into a project and keep everything neat. Adding APIs that would allow us to have inline UXML/USS would be greatly appreciated.

  • Generally expand the public API of UITk. Right now, the design of UITk is very extensible but a lot of that extensibility is locked away behind internal. Would be great if Unity could make an effort to solidify those APIs and make them public.

2 Likes

Been using UI Toolkit about a month or so, only runtime. Right now, once I understood the design philosophy of the package the things i’m missing most are below. I should mention my use case are not exactly games, we do 3D apps for industry and as such my expectations for the UI Toolkit are to be similar to other “desktop apps”, like webs or XAML, so I would like more priorization for stability than the fancy stuff like custom shaders (not that we don’t have use for them too, but developer experience right now is a little rough and may be more important for adpting UITK).

UI Builder QoL

Better workflow for USS and UXML

Right now (unity 6.2) it feels very rough for runtime UI. Some things are exclusive text-mode on USS or UXML (Images, TwoPaneSplitView…). Not a big problem by itself, except for discoverability, as these things are not on the ui components library.
But… it is a pain to switch between UI Builder and text. Builder’s design is for full screen use, so you have no direct access to the Project tab. [And if you have to open the UXML on text mode, unity editor doesn’t open it on double click or have a context menu option to do it, you have to explore the folder and then open it on your text editor.] >> Just discovered while writing there is an Open with IDE button on the Project tab of the library for the UXML. But the workflow for USS still depends on the main editor projects tab.
The best solution for this should be make it editable on the builder itself, making the “Show UXML/USS preview” a fully fledged code editor with syntax highlight like Visual Studio XAML editor. But if that is not easy, at least a “open in text editor” button somewhere would make things softer.

Preview datasets for ListView, TreeView etc

Right now designing with ListViews and other collection-based components in the UI builder is imposible. You have to run the app and navigate to that UI just to see it has a wrong margin or something. Having the listview/treeview just spawn a configurable number of elements with the given Item Template would make it much better.

UI Builder panel sync with Game View / Simulator

For WYSIWYG design it is very important to have a way to set the UI Builder’s panel settings like runtime. For mobile you would design with constant physical size and/or UI scale, specially if you want to deploy to both mobile and pc or phones and tablets. And given the Builder uses a panel as can be seen on the debugger, it doesn’t look too difficult to implement, just copy the given panel asset’s settings.

More code automations and/or better debugging for UXML/USS

UXML having text-based references is a big problem for any kind of refactor. It is what it is and in web world happens the same with Javascript. But on unity some things could be improved, for example, after moving around some files there are warnings about files moved. On files with GUID this warning should fix itself, if the code already knows it has moved, just keep the UXML updated.
Debugging has its own problems, each imported file throws it’s own bugs but there is no reference on click to the failing UXML or USS, nor in the text itself. For example, after adding a namespace to some custom components. Once you have a few different UXML files it becomes a pain in the ass to just search by hand for the ones using the changed reference. Just adding the context to the console so it can point to the file, or adding the file path in the error text, would be a great improvement.

Under development features

I think the most important ones right now, after we have World Space UI that was very important for VR, are having better CSS parity. Media Queries, Z-index and layout based pseudo classes are a must. Grid layout would be great too, and box-shadow has a lot of uses for ui separation and adding visual depth to prompts or indent cards.

5 Likes

Gap, additional pseudo-classes and more advanced, customizable binding to achieve better performance for heavyweight interfaces would be great news for me. I often use containers with elements inside, which sometimes are problematic to set up in terms of margin and padding, so Gap would help a lot with that, as well as pseudo-classes.

2 Likes

I would like improvements on buttons and navigation.

I want the ability to bind buttons to functions directly from the UI Builder the same way you can bind other data.

I want the ability to customize the navigation between elements directly from the UI Builder.

I want the ability to directly bind input actions to press buttons in the UI Builder.

Other then that from what’s already on the list, I want grid layouts and addressable support.

1 Like

That can save so much time, I can’t even imagine.

Oh yeah

2 Likes