Clarifications
This page serves as a correction to false information and rumors surrounding the project.
The aim is to clarify any misconceptions and provide accurate information to keep everyone informed.
This page doesn't contain any technical information related to the project, You can skip it if you're not interested.
Jenova and AI
With the rapid growth of Artificial Intelligence (AI) and, in particular, Generative AI in recent times, it has become increasingly challenging to find a project that doesn't utilize AI in some way. However, some individuals often use AI-related accusations to tarnish a project's reputation or manipulate public perception against it. In this section, I aim to provide transparency by openly sharing how and where AI has been employed within the project.
AI Slop Labeling
Some ʺinternet journalistsʺ have unfairly labeled the project as ʺAI Slopʺ solely because of the existence of an addon called JenovaLLM.
This is a misconception. Projekt Jenova is not inherently focused on AI and does not promote AI in any way. Instead, its primary purpose is to push the boundaries of the engine and introduce innovative technologies.
It's worth noting that Large Language Models (LLMs) are a relatively new and cutting-edge technology and JenovaLLM is merely an addon within the framework, just like many others. Thus, it has no direct connection to the framework's core objectives.
Why JenovaLLM Exists?
Jenova Large Language Model (JenovaLLM) is an addon specifically designed for game developers who want to incorporate intelligent virtual characters or complex game logic using a language model. However, it's essential to note that the addon only serves as an inference engine powered by llama.cpp backend for various model formats, without providing any pre-trained models or data. This allows developers to utilize both large and small language models stored as GGUF files in real-time, leveraging the power of Vulkan compute kernels.
I developed the addon during my research and studies on large language models, focusing on technical production aspects. While it can be used as a chatbot, its potential extends to various other usecases. Essentially, it's a versatile tool that can be employed in multiple contexts, much like any other utility in the developer's toolset.
Vibe Coding Accusations
Few individuals have accused Projekt Jenova of ʺVibe Codingʺ without providing any concrete evidence. To be honest, I was quite surprised by this claim. I began developing the project at the same time I was learning the engine and my involvement in the project is fully transparent.
Jenova Framework is a product of approximately 10+ years of experimental development. It began with the name Morpheus++ in 2015 as a C++ scripting framework for Autodesk 3ds Max. At that time, ʺVibe Codingʺ did not exist. Over the years, it evolved as I gained more experience, and it was eventually recreated from scratch for the Godot Engine in 2024.
The codebase is entirely human-written. I encourage anyone interested to review the source code for verification.
In fact, the Jenova Runtime is one of the most complex yet clean and organized codebases ever created, If someone believes that this level of complexity could be achieved by an existing AI language model, they are sadly mistaken. These unfounded accusations served to undermine the project's value and you can learn more about these claims at the Jenova vs Godot section.
Akira Mascot Logo
Before developing Projekt J.E.N.O.V.A, I spent a year on an project called MaxFusion which was an in-house generative AI developed in C++ for Autodesk 3ds Max by providing a layer-based post-processing toolkit powered by diffusion technologies.
MaxFusion models were trained entirely on my own portfolio and hundreds of images I personally created. The initial logo of the project was created based on an AI-generated image from MaxFusion. While it was a placeholder, it wasn't a raw-generated image and it had many layers of repaint over it for better details.
As of 2026, I redesigned the logo and it's been replaced with a fully Computer-Generated Image (CGI) made in 3ds Max and ZBrush.
Where AI is Used
Here's a clear list of where AI is used in the project ecosystem:
-
Text Refinements: While all documentation and pages are written by myself with precision, Llama3-8B-Instruct model is used to refine the grammar of GitHub and Documentation pages with a slight edit. None of the texts is written directly by language models.
-
Announcement Banners: Some early announcements banners on Discord featured AI-generated images with hand-painted refinements. However as of 2026, all banners will be fully CGI similar to the logo.
-
Technical Research: Sometimes AI is used to gather technical information and academic papers for some components.
-
Codebase Analysis: Time to time, advanced deep-learning analyses are performed on the codebase to visualize the logical flow.
-
Upscaling and Rendering: JenovaRTX utilizes DLSS 4.0 which employs AI-powered technologies in upscaling and ray reconstruction.
Jenova vs Godot
Since Projekt J.E.N.O.V.A entered the scene, Godot community have been... unwelcoming.
It was never formally acknowledged and parts of the community attempted to belittle it, labeling it a toy, scam, useless, ʺjust a wrapperʺ etc.
At this section, I address the false claims and the history that transpired between the project and Godot community.
This section presents a narrative of interactions within the Godot community based on personal experience.
It is not a comprehensive record and does not represent the views of the entire community.
Scam Accusations
When I introduced the project to the community for the first time, many people attacked the project and called it a scam, basing their argument on ʺC++ scripting isn't real, It can't exist!ʺ They were dismissing the capability of the project and me the developer.
When I provided the proof-of-concepts, the response was ʺOpen-source it or you're lying!ʺ So I did...
The same pattern of negativity and skepticism repeated itself when I released screenshots of JenovaRTX showcasing real-time path-tracing inside the engine. Many community members called me a liar, a joker, a faker and a scammer, despite me not selling anything.
After providing videos and technical details of the project, the critics vanished, No apology...
Rejection & Denial
After a while, the project received attention from people who were genuinely interested in its capabilities and direction.
However, some community members, even those who used the framework, spread misinformation and misconceptions about the project.
Most repeated claims included calling the project ʺJust a GDExtension wrapperʺ or ʺHaving almost no documentationʺ
-
Claim Check: Is Jenova Framework Only A GDExtension Wrapper?
The answer is a definitive NO, Jenova Runtime is a complex ecosystem and features a fully-featured scripting backend with nearly all the features presented in GDScript. It has the most comprehensive feature list after GDScript, standing above all other third-party languages for the engine. Hours of reverse-engineering GDScript were invested to implement its features in the framework.
Jenova only utilizes GDExtension for communication with the engine, While it has the capability of wrapping GDExtension, It provides a complete scripting backend including an Advanced Interpreter, C++ Script Instance, C++ Script Object, C++ Compiler Interface, Metadata Extractor, C++ Language Implementations, Property Mechanism, Built-in Profiler, Package Manager, Addon/Tool System and so on.
Nobody would make a statement like that if they've even read a single block of the project's source code :D -
Claim Check: Does Jenova Framework has Almost No Documentation?
Jenova is a complex project with numerous features and documentation has been an ongoing process since its public release. While some features may still be missing from the documentation, there is comprehensive coverage for the most part of the framework. I've invested a considerable time in creating a dedicated, easy-to-use and well-designed documentation and learning center, perfection takes time.
FAQ
Who is behind Projekt Jenova?
Projekt Jenova is designed, developed and maintained by a single individual — me.
Why is it called Projekt J.E.N.O.V.A/Jenova?
The logic behind the naming convention is detailed in the Introduction, but in short, "Godot" can be translated to "Small God" and "Jenova" to "New God," representing a new, more powerful god emerging from a smaller one.
Is Projekt Jenova Open-Source or only part of it?
Jenova Runtime (the heart of the framework) itself is fully open‑source under the MIT license. Some optional technologies built using the framework such as JenovaFibers, JenovaRTX, JenovaIDE, JenovaFusion or other in‑house components are proprietary because they were developed specifically for my own game.
The naming of components like JenovaRTX indicates they are built within the framework, not that they are simply part of the project.
Can I work with other languages or is it only for C++?
Yes, while Jenova Framework primarily focuses on C++, it also provides support for languages like C, Clektron and Blade for more lightweight tasks.
Which compiler works best with Jenova?
Microsoft Visual C++ (MSVC) is the most performant and compatible compiler for the Jenova Ecosystem.
What are future plans for this project?
The future is currently undefined. The Framework is in a strong position and feature-rich, but its development will depend on the needs of my game during its development stages.
How can I support your work?
Send me one billion dollars!