My thoughts on UE6 ditching Blueprint

// 06.18.2026

I'm seeing a lot of takes with UE6 ditching Blueprint. My take? THANK GOD and good riddance. After working on and shipping a UE5 game for over two years, I genuinely believe that any team would be better off in the long run learning C++ from the start. I am saying this as someone who started with little C++ or BP knowledge and learned BP first before getting more comfortable with C++. Blueprint is more accessible, definitely, but also has a very low skill and complexity ceiling compared to C++ and quickly becomes the worst tech debt ever if you hit those ceilings. And when you do, because no Blueprint-declared variables are accessible in C++, you will spend weeks or even months rewriting and redeclaring everything in C++ just to be able to access the variables that are locked away. It becomes a domino chain of inaccessible dependencies. This is coming from first-hand experience.

You can't convince me that constantly organizing nodes and clicking through multiple menus just to define a struct is more time efficient than typing out a few lines in C++. Blueprints are unmaintainable, often much harder to understand and read, and harder to debug. I feel like the "math expression" node is a perfect example of this. If it's that much easier to type out and read a math expression than manually dragging every operand and variable, connecting and organizing them, then why not just type your whole program out anyway???

Games are also inherently complex. Not a beginner's programming project if it is to be done right. You are running a world simulation loop that should ideally finish consistently at least 60 times per second with no hitches. You need to know the size and time cost of everything that you do. You need to be running things async. I believe Blueprint is abstracting too much and abstracting things that are inherently too complex to be abstracted.

In addition to all of this, love it or hate it, AI has had a notable increase in my productivity, allowing me to offload the syntax and implementation details to save my brain for the higher-level architectural problems. Text is accessible to AI. BP is not. In such a competitive industry, it's hard to imagine leaving those productivity gains on the table at this point.

Obligatory Expedition 33 acknowledgement given that it is the 100% Blueprint game everyone points to: The game runs like shit on Steam Deck. I'm pretty confident the CPU would run better if it wasn't 100% Blueprint. Also, obviously, default UE5 rendering, but that's another discussion.

Now, Verse. What is supposed to replace C++ and BP. I have never used it, so it's honestly too early to say. We will have to see in the long run. But I like what I see so far. Async by default and functional? Heavy reliance on the single-core game thread is one of the biggest performance limitations of UE5 today. Functional programming is also an objective win for maintainability and code safety. I will say that the small amount of syntax I've seen so far felt a bit weird and too complex. I will acknowledge that the lower-level nature of C++ leaves a ton of foot guns and some additional complexity, so if Verse can strike a healthy middle ground, I think it could be good.

Addendum

I think visual scripting still has a place for simple exposed cosmetic events for artists to hook onto. Also, for programming shaders, as that is a highly visual type of programming.