VoxelModel-v1: a new open-source text-to-3d model on HuggingFace

The model bench-labs/VoxelModel-v1, which solves the text-to-3d task, has been published on HuggingFace. The model appeared on July 26, 2026, and the code has been released as open source. The developer does not specify VRAM requirements, export formats, or other technical details.

What it means

VoxelModel-v1 is a 3D generator from a text description. Among the tools in our reference, there are currently no direct open-source equivalents for the text-to-3d task. The closest related generators from the AI 3D section — OpenLRM, Hunyuan3D 2.1, and TRELLIS — work on the image-to-3d task, meaning they require an image as input rather than text. Hunyuan3D 2.1 mentions text-to-3d in its repository tags, but its main task is still image-to-3d. Since the developer of VoxelModel-v1 does not disclose technical specifications, it is not yet possible to compare the new model with the listed tools by VRAM, model weights sizes, or export formats. The model has just appeared, and there is little detail about its capabilities. If the text-to-3d task is relevant to you, it is worth following updates — examples of its work and technical documentation may appear soon.