Hunyuan3D vs InstantMesh

Hunyuan3D is Tencent’s system for generating 3D assets from text, images, or sketches. The international platform accepts up to four multi-view images, offers model editing and adjustment, and exports OBJ and GLB files. Its capabilities include text-to-3D, image-to-shape, and texture generation; Smart Topology supports triangular and quadrilateral faces. Tencent Cloud documents API tasks for generation, texturing, topology reduction, component generation, UV mapping, motion, auto-rigging, and format conversion. Hunyuan3D-2.1 releases full model weights and training code, and its repository documents a local Gradio app. A Blender add-on is listed for Hunyuan3D-2. The stated VRAM needs are 10 GB for shape generation, 21 GB for texture generation, and 29 GB for both. First-time API activation grants 100 credits, valid for one year. Paid packs include 1,000 credits for 100 CNY, paid once and 10,000 for 980 CNY, paid once; pay-as-you-go is billed at 0.12 yuan per credit, with monthly settlement.

InstantMesh is a free, Apache-licensed framework that generates a 3D mesh from one image using the LRM/Instant3D architecture. It runs on Linux in a local or self-hosted setup. A Gradio demo starts with `python app.py`, while `run.py` offers command-line generation. The default export is an OBJ mesh with vertex colors; an alternative adds a texture map and takes longer. rembg can segment the foreground object, or that step can be skipped when the source already has an alpha mask. Four sparse-view reconstruction model variants and a customized Zero123++ UNet for white-background image generation are included. Docker deployment is documented. The recommended environment uses Python 3.10 or later, PyTorch 2.1.0 or later and CUDA 12.1 or later; two GPUs can reduce demo memory use. Training code is included, but the training dataset is unavailable because of its size. Community implementations are noted for ComfyUI, Replicate and Google Colab.