Leaderboard / Model

Trellis 3D

by Microsoft3D generatorOpen source, free to run locally

Microsoft's open-source image-to-3D model, a favorite for local generation pipelines.

How to use Trellis 3D with Blender

Run TRELLIS locally or through a hosted space, export the mesh as GLB, and import into Blender. Popular in ComfyUI-to-Blender pipelines.

Best for: Developers and researchers who want state-of-the-art image-to-3D without usage costs.

Which AI built it better?

Two random Season 3 entrants answered the same prompt. Pick the better build and the names are revealed. Your vote counts toward the leaderboard.

Picking a matchup…

AModel A
BModel B

Not benchmarked yet. Trellis 3D has no entrant in Season 3, so there are no builds, ratings or votes to show. How tools enter the arena

What is Trellis 3D?

TRELLIS is Microsoft Research's open-source 3D generation model. It converts images (and text via image conditioning) into 3D using a unified latent representation that can decode to meshes, radiance fields, or 3D Gaussians.

Since release it has become the default open-source image-to-3D baseline — widely run in ComfyUI, Hugging Face spaces, and research projects. Quality on single-image inputs rivals commercial services, which is remarkable for a free local model.

Strengths and limitations

Strengths

  • Strength:Open source and free — unlimited local generation
  • Strength:Near-commercial quality image-to-3D
  • Strength:Flexible outputs: mesh, radiance field, or Gaussians
  • Strength:Large community: ComfyUI nodes, HF spaces, forks

Limitations

  • Limitation:Requires a serious GPU (16GB+ VRAM for comfortable use)
  • Limitation:No official web app — setup is on you
  • Limitation:Texture quality trails Hunyuan3D and commercial tools