NC AI said on the 1st that it will unveil the three-dimensional (3D) generative artificial intelligence (AI) service "VARCO 3D 2.0." With this update, the company plans to apply AI across the entire real-world content production process and expand its use into industrial fields such as physical AI and Digital Twin.
VARCO 3D is the only 3D generative AI service in Korea, and the 2.0 version is characterized by streamlining repetitive tasks on production sites and enabling workers to directly design and control the production process.
The core of this update is the "Custom Workflow" feature, which lets users freely combine various functions—image generation, AI assistant, 3D generation, texture conversion, and post-processing—at the node level to design a tailored production process that fits project goals. Complex, repetitive production processes in the field can be saved as a single workflow and reused whenever needed.
Saved workflows are used to quickly create multiple 3D design drafts in the early concept planning stage. Even if the planning direction changes or feedback arises, revisions can be reflected based on the existing workflow.
They can also be applied to work that builds characters or objects by dividing them into parts such as clothing, accessories, and equipment. This reduces shape distortion that can occur when generating multiple elements simultaneously and allows individual edits to specific parts, easing the burden of having to regenerate the entire model.
Team-based collaboration is also expected to get easier. Completed workflows can be shared with team members, allowing a production pipeline that has passed field validation to be applied uniformly across projects. The company said this internalizes individual production know-how as a shared organizational asset and provides an optimal collaboration environment in which many workers produce high-quality 3D assets under consistent standards.
According to NC AI, VARCO 3D posted strong performance on major benchmarks, including "Uni3D," a shape similarity evaluation metric. Compared with global 3D generative AI services such as Hunyuan, Meshy, and Tripo, it showed competitiveness in original shape preservation, fine detail, and texture quality.
NC AI plans to expand its enhanced 3D generation performance and Custom Workflow features into the fields of physical AI and Digital Twin. The company expects they can be used to produce industrial assets that require fine specification changes—such as factory equipment or robot parts—by module and to mass-produce them under consistent standards.
In the third quarter of this year, the "VARCO 3D 2.0 Ultra" model is set to be unveiled. The model is expected to maximize the realism of characters and objects by precisely realizing micro shapes that existing generation models found difficult to reproduce, such as wrinkles, skin contours, and surface irregularities.
NC AI CEO Lee Yeon-su said, "This VARCO 3D 2.0 update goes beyond simple Generative AI and focuses on innovating practitioners' 3D content production pipelines themselves," adding, "We will continue to set the standard for high-quality 3D asset production based on overwhelming AI technology and expand applications into various industrial fields, including physical AI and Digital Twin."