I was working on a video project recently and really wanted to add some simple 3D text to a scene. I realized that doing this natively in Shotcut is currently very challenging and nearly impossible without using external software and importing transparent clips.
This got me curious, so I did some searching online to see how other software handles it. I found out that since Shotcut is built on Qt, there is a module called Qt Quick 3D. From what I’ve read, it’s designed to bring 3D objects into a 2D interface without needing a massive overhaul of the engine.
I’m not a developer, so I don’t know the exact complexity of this, but I think it could be a game-changer if we could eventually:
Add 3D Text: Have a filter where we can rotate and scale text in a 3D space.
Import Simple 3D Models: Even something simple like being able to drop in a modeled car or a basic shape (glTF/OBJ) and move it around the screen would be incredible for creators.
I think even a very “basic” version of this—just the ability to tweak the position and rotation of a 3D object—would put Shotcut on a whole new level compared to other free editors in 2026.
I’d love to hear from the developers or anyone more technical—is something like Qt Quick 3D a possibility for the long-term roadmap, or is it much more complicated than it looks?
Thanks for all the hard work on this amazing editor, currently working and testing the new stable 25.12.31 version of the software
Qt Quick cannot be used in the engine due to its threading model. It is mainly a UI technology.
It might be possible to use it like the HTML generator where Shotcut drives its UI to generate an image or video. However, it is rather naive to assume one can simply type in text and choose an extrusion depth or import a model and be done. There is so much more to 3D. This is basically opening a door to a whole new world and tool. It is an invitation for massive feature creep.
I did a search for Qt 3D-based tools, and there is basically none other than Qt’s own Design Studio, which is more intended for UI app development. Maybe the reason for this is the quality of Blender. And if some code-oriented UI were presented like the HTML generator, it is almost useless due to the lack of knowledge in this area even with the help of AI. Most often the AI only gets you close, but you still need to do manual work or get frustrated iterating with prompts. The HTML version works because of the sheer volume and maturity of web content combined with the massive community around codepen.io
Maybe the best thing to do here is to try to get WebGL working in the HTML generator.
OpenShot has Blender integration that is similar to this generator approach. That could possibly be done as well, but you can also simply use OpenShot for that purpose right now. I have heard of people using it to generate a title to use within Shotcut. I have also read about some versioning and compatibility issues. It might be a bit sensitive in that regard because it is not based on a rather stable protocol like Chrome DevTools that our HTML generator uses. Instead, it uses the blender Python API.
1 Like
Thanks for the detailed technical explanation, Dan. The threading model conflict with Qt Quick makes perfect sense. I hadn’t considered the friction between the MLT engine and the UI framework.
Your point about WebGL in the HTML generator seems like a very pragmatic middle ground. It avoids ‘reinventing the wheel’ regarding 3D engines while leveraging the massive amount of existing web-based 3D content and libraries.