Article

Roblox launches Cube 3D: an open AI model for creating 3D objects

Picture of Sheet

Sheet

time: 4 min read

Roblox has announced the launch of Cube 3D, an open AI model that allows developers to create 3D objects using text commands, with the aim of speeding up the development process and expanding the variety of content on the platform.
The 3D creation revolution is at hand

On March 17, 2025, Roblox announced the launch of Cube 3D, an innovative open-source AI model designed to make creating 3D objects more accessible and faster than ever before. The model allows developers and designers to create “meshes” (3D representations of objects) by typing simple text commands.

“With Cube 3D, all a developer has to do is type something like ‘create an orange race car with black stripes,’ and the model automatically creates a high-quality 3D object that can be immediately incorporated into a game,” explained David Bezucki, CTO of Roblox. “This is a paradigm shift in how people create 3D content.”

One of the significant advantages of the new system is its full integration with Roblox Studio, allowing developers to further enhance objects created using the platform’s familiar tools. This hybrid approach allows developers to take advantage of AI technology while maintaining full control over the creation process.

Technological innovation with a unique approach to 3D

What sets Cube 3D apart from similar solutions on the market is its innovative technological approach. Rather than relying on traditional approaches to creating 3D models, Roblox developed a unique tokenization technique, in which 3D objects are represented as a sequence of tokens – similar to how advanced language models represent text.

“The model is trained to predict the next token in the sequence to build a complete 3D object,” explained Dr. Susan Chang, a senior research scientist on the development team. “This approach allows Cube 3D to rapidly create functional and accurate 3D models, specifically tailored for game engines and the performance requirements of interactive platforms.”

Another significant advantage is that the model is developed as open source, allowing the global developer community to customize, create extensions, or train the model with unique datasets. This move represents a strategic shift for Roblox, which has traditionally tended to keep core technologies as closed intellectual property.

“We decided that the value of the ecosystem that would be created around Cube 3D was greater than the competitive advantage of keeping the technology to ourselves,” noted David Bezzochi. “We believe that the developer community will take the foundation we have created and expand it in directions we have not even thought of.”

 

Revolutionary impact on the development and creation process

The new tool is expected to dramatically speed up the development process on the Roblox platform, allowing developers to explore new creative directions more quickly and efficiently than previously possible. According to the company, creating a high-quality 3D object that could previously take hours or even days can now be done in minutes.

“The model is designed to remove the biggest barrier for new creators – the need for advanced 3D design skills,” explained Emily Chen, senior product manager at Roblox. “We want to open up the world of 3D creation to the millions of people who have great ideas but lack the technical skills to implement them.”

The move fits into Roblox’s broader strategy to expand its audience beyond traditional gamers and transform the platform into a multi-purpose virtual space. The availability of 3D creation tools is expected to attract developers from fields as diverse as education, architecture, product design and digital art.

Industry analysts estimate that the move could increase the amount of content created on the Roblox platform by 30% to 50% over the next year, while expanding the range of experiences available to users.

 

Looking to the future: Plans to expand capabilities

Roblox is already planning the next steps in Cube 3D’s evolution, with a variety of improvements and new features in the pipeline. “We’re working on expanding the model’s capabilities to accept images as input,” Chen revealed. “The goal is to make Cube 3D a true multimodal model, capable of combining different types of input such as text, images, and even video.”

Additionally, the company is working on deeper integration between Cube 3D and existing Roblox Studio creation tools, including the platform’s physics, animation, and lighting systems. This integration will allow developers to create richer, more interactive experiences, with a focus on performance and optimization.

“Cube 3D is just the first in a series of AI tools we plan to launch in the coming year,” promised Bazuki. “We are working on a variety of solutions, including text generation, text-to-speech, and speech-to-text tools, that together will provide developers with a comprehensive toolbox to turn their ideas into virtual reality.”

 

Impact on the gaming industry and meta-warfare

The launch of Cube 3D represents a broader trend in the gaming industry and the metaverse, where AI technologies are becoming central tools in the creative process. Roblox’s move is a direct response to the competitive challenge from companies like Epic Games (maker of Fortnite and Unreal Engine) and others, who are also investing significant resources in AI-powered content creation tools.

“The battle for the future of the metaverse will be decided not just by who has the best technology, but by who can build the largest and most diverse creative community,” said John Raccio, senior analyst at research firm Gartner. “Roblox’s move to open up their technology to the broader community puts them in a strong strategic position in this competition.”

Summary: A new era of democratization in 3D content creation

The launch of Cube 3D marks a significant step toward a future where 3D content creation is accessible to anyone, regardless of technical background or advanced design skills. The move aligns with Roblox’s long-term vision to create a platform where anyone can build, share, and experience almost anything imaginable.

“We believe that the virtual world of the future will be built not by large corporations, but by millions of independent creators,” concluded David Bezucci. “Cube 3D is a critical step in empowering our community of creators and expanding it beyond traditional game developers.”

With the continued growth of virtual spaces and the growing interest in metaverse, Roblox’s ability to continue to innovate in the area of creative tools could be a crucial factor in maintaining its leading position in this competitive and dynamic industry.

Picture of Sheet

Sheet

Morbi art orci, malsuada sd fulviner in, dictum et risus. Elicum art volutpat. Nam eros lactus, tempor quis condimentum nec, viverra ac mi. Vivamus id ex sd quam agustas uimod a sit amet dolor. Morris elementum audio a niva conga socifit.

Picture of גיליון

גיליון

מורבי ארט אורצי, מאלסואדה סד פולווינר אין, דיקטום את ריסוס. אליקום ארט וולוטפט. נם ארוס לקטוס, טמפור קוויס קונדימנטום נק, ויווררה אק מי. ויוומוס איד אקס סד קואם אגסטס אוימוד א סיט אמט דולור. מוריס אלמנטום אודיו א ניבה קונגה סוסיפיט.

Sign up now

You will receive the latest news and updates about your favorite categories!

Related article

A team of Harvard researchers introduces TxAgent: an artificial intelligence agent for personalized care

 

TxAgent: A Breakthrough in Personalized Medical Care

The Zitnik Lab at Harvard University recently introduced TxAgent, an innovative artificial intelligence agent designed to revolutionize the therapeutic decision-making process. The development comes at a critical time as healthcare systems around the world seek ways to improve the efficiency of medical care and reduce medical errors.

Led by Prof. Marinka Zitnik, the team of researchers developed a system that combines advanced multi-step analysis and real-time dynamic access to huge medical databases. The system is able to assess complex drug interactions, identify potential contraindications, and suggest treatment strategies tailored to the specific patient based on their entire medical data.

“TxAgent represents a completely new approach to precision medicine,” explains Prof. Zitnik. “Instead of relying on standard protocols that are not always appropriate for every patient, our system analyzes each patient’s unique situation and offers a personalized treatment plan.”

TxAgent’s complex architecture integrates 211 different data analysis tools, ranging from traditional statistical models to advanced deep learning algorithms, all working in harmony to provide accurate and reliable treatment recommendations.

 

Advanced capabilities and technological innovation

One of TxAgent’s significant advantages is its ability to perform complex analyses of medical data in real time. The system can analyze a patient’s medical history, laboratory test results, genetic data, and information about previous treatments, and combine all of this with the latest scientific knowledge from global medical databases.

Unlike traditional clinical support systems, TxAgent goes beyond providing basic alerts, but offers in-depth, layered analysis:

  • Drug interaction analysis : The system identifies not only direct interactions between drugs, but also combined effects of several drugs together.
  • Genetic matching : TxAgent takes into account the patient’s genetic profile to predict drug efficacy and side effects.
  • Continuous learning : The system is constantly improving with each new case, updated in real time from new studies and clinical protocols.
  • Transparency in decision-making : Unlike “black box” systems, TxAgent provides clear explanations for each recommendation, which increases the medical team’s trust in the system.

Dr. James Chen, one of the senior researchers on the project, emphasizes: “One of the biggest challenges in medicine today is the enormous amount of information that doctors have to process. TxAgent helps filter the most relevant information and presents it clearly, allowing doctors to make more informed decisions in less time.”

 

The potential to improve treatment outcomes and patient safety

The clinical application of TxAgent holds significant potential for improving treatment outcomes and patient safety in several aspects:

  1. Reduce medical errors : According to studies, medication errors cause many unnecessary hospitalizations each year. TxAgent can reduce these errors by accurately identifying drug interactions and contraindications.
  2. Dose personalization : The system is able to recommend dosage adjustments based on many factors, including patient weight, kidney and liver function, medical history, and genetic profile.
  3. Improving treatment compliance : Through personalized recommendations, TxAgent can help reduce side effects and improve patients’ compliance with drug treatment.
  4. Streamlining decision-making processes : Making relevant information available in real time allows the medical team to make faster and more informed decisions.

In initial trials in a controlled clinical setting, TxAgent showed a 28% improvement in identifying potentially dangerous drug interactions, and a 34% improvement in tailoring treatment protocols to unique patient profiles.

 

Ethical challenges and questions

Despite the great promise, the development and implementation of systems like TxAgent also raises significant challenges:

  • Privacy and data security : The system requires access to sensitive medical information, which raises questions about data security and privacy.
  • Medical liability : Questions regarding liability in cases of error or system failure have not yet been fully resolved.
  • Integration into existing systems : Integrating the system into existing medical information systems poses a significant technical challenge.
  • Training of medical teams : The success of the system depends largely on the ability of medical teams to use it effectively.

Prof. Zitnik is aware of these challenges: “We take the ethical and practical questions seriously. TxAgent is intended to be a decision-support tool, not a substitute for medical judgment. The physician always remains the ultimate authority in the decision-making process.”

Summary: The Future of Personalized Medicine

TxAgent represents a significant step forward in the era of personalized medicine. The system demonstrates how artificial intelligence can be used as a powerful tool in the hands of medical staff, enabling more informed decisions tailored to the unique needs of each patient.

“Personalized medicine is not just a trend, but the future of medicine,” concludes Prof. Zitnik. “With tools like TxAgent, we are moving toward a world where medical care will be more precise, more efficient, and perfectly tailored to each patient.”

The research team is now planning larger clinical trials in collaboration with several hospitals across the United States, with the aim of testing the system’s effectiveness in a wide range of clinical scenarios and patient populations. If the trials are successful, TxAgent could be available to medical institutions within two to three years.

We tailor unique financing solutions for importers based on deep knowledge of the business.