这些 3D 生成器免费吗?
这些模型页面在抓取时是可公开访问的项目。具体是否排队、休眠、限流或改变可用性,取决于每个模型页面的作者和运行配置。
测试已检查的公开 3D 模型队列,覆盖图片转 3D、文字转 3D、网格生成、Gaussian splat、深度工具和 3D 资产实验。暂停、报错或不可访问的模型会从目录里移除。
tencent/Hunyuan3D-2
Text-to-3D and Image-to-3D Generation
microsoft/TRELLIS.2
High-fidelity 3D Generation from images
ilcve21/Sparc3D
Next-Gen High-Resolution 3D Model Generation
TencentARC/InstantMesh
Create a 3D model from an image in 10 seconds!
stabilityai/stable-fast-3d
Stable Fast 3D
VAST-AI/TripoSG
TripoSG
stabilityai/TripoSR
TripoSR
trellis-community/TRELLIS
Scalable and Versatile 3D Generation from images
hysts/Shap-E
text-to-3D & image-to-3D
depth-anything/depth-anything-3
Depth Anything 3
tencent/Hunyuan3D-2mv
MultiImages-to-3D Generation
TencentARC/Pixal3D
High-fidelity pixel-aligned image-to-3D generation.
VAST-AI/TripoSplat
TripoSplat
frogleo/Image-to-3D
Speedy and Accurate Image to 3D Generator
fffiloni/PSHuman
PHOTOREALISTIC HUMAN RECONSTRUCTION w/ CROSS-SCALE DIFF
tencent/HunyuanWorld-Mirror
Universal 3D World Reconstruction with Any Prior Prompting
Roblox/cube3d-interactive
interactive demo for cube 3d model
webml-community/bonsai-ternary-webgpu
Run 1.58-bit Bonsai LLMs locally in your browser on WebGPU
sudo-ai/MeshLRM
MeshLRM (Unofficial)
VAST-AI/MV-Adapter-Img2Texture
Generate 3D texture from image
facebook/map-anything
Mapanything Gradio
facebook/ActionMesh
Generate animated 3D meshes from video
VAST-AI/MV-Adapter-Text2Texture
Generate 3D texture from texts
3DTopia/3DGen-Arena
3DGen Arena
gagndeep/Apple-Sharp-Image-to-3D-View-Synthesis
Convert any static image into a 3D Gaussian Splat scene
alexnasa/pixel3dmm
Versatile Single-Image 3D Face Reconstruction
gangweix/Pixel-Perfect-Depth
pixel-perfect monocular depth estimation
Ruicheng/MoGe-2
Monocular metric-scale geometry estimation
TencentARC/GeometryCrafter
GeometryCrafter
Ruicheng/MoGe
MoGe live demo
tencent/Hunyuan3D-Part
Hunyuan3D Part
VAST-AI/AniGen
Generate animatable and articulated 3D assets from images
VAST-AI/DetailGen3D
DetailGen3D
CompVis/EDGS
EDGS
facebook/vggt-omega
3D reconstruction from images/video with VGGT-Omega
pablovela5620/mini-dust3r
Mini Dust3r
A19grey/Depth-Pro-in-Meters
Depth-Pro-in-Meters
kxic/EscherNet
3D novel view synthesis from any number images!
HorizonRobotics/EmbodiedGen-Image-to-3D
Generate physically plausible 3D model from single image.
prithivMLmods/TRELLIS.2-Text-to-3D
Text-to-Image → 3D or Image-to-3D
sudo-ai/SpaRP
Generate 3D mesh from unposed images | Hillbot
cakewalk/splat
WebGL Gaussian Splat Viewer
lpiccinelli/UniK3D-demo
UniK3D (CVPR 2025)
imlixinyang/FlashWorld-Demo-Spark
Spark frontend for FlashWorld
multimodalart/HunyuanWorld-viewer
HunyuanWorld Viewer
AleenDG/3DGenTripoSR
TripoSR
Asrar990/3D_Reconstruction_Tool_From_Video
3D model reconstruction from video
prs-eth/depth-to-3d-print
Depth To 3D Print
webml-community/TinySwallow-1.5B-Instruct-WebGPU
A compact Japanese LLM that runs locally in your browser.
VAST-AI/TripoSG-scribble
TripoSG Scribble
merve/daggr-image-to-3d
Daggr Image To 3d
victor/pixal3d-studio
Clean image-to-3D studio for Pixal3D
cyun9286/Align3R
Aligned Monocular Depth Estimation for Dynamic Videos
victor/fable-5-boeing-747
Procedural Boeing 747-400 in Three.js, built by Fable 5
prs-eth/PaGeR
Panorama Geometry Reconstruction
Insta360-Research/UniSHARP
Single-image 3D Gaussian novel view synthesis.
quanghuynt14/fable-5-watch-movement
Mechanical watch movement in Three.js, built by Fable 5
haoz19/world-tracing-demo
Multilayer-geometry 3D from a single image or 16-frame clip
huggingface-projects/diffusiongemma-3d-gen
Diffuse an SVG game asset and watch it extrude into live 3D
Roblox/cubepart-demo
Open-vocabulary part-controllable 3D mesh generation
developy/ApDepth
ApDepth Official Demo
karimknaebel/surge
3D reconstruction from images with SurGe
openfree/drqxab
3D Simulation
build-small-hackathon/karate-wiener
Create your own karate moves with a sausage man
victor/opus-4.8-boeing-747
Opus 4.8 Boeing 747
bartduis/modality_forcing
Text → RGB + depth + 3D point cloud.
这些模型页面在抓取时是可公开访问的项目。具体是否排队、休眠、限流或改变可用性,取决于每个模型页面的作者和运行配置。
很多公开队列不需要 Photo3D 账号即可测试。公开 GPU 忙时仍可能限流、休眠或需要排队等待。
建议先当作测试和研究工具使用。正式使用前要检查拓扑、比例、贴图、授权和后续软件兼容性。
公开模型页面可能会暂停、构建失败、运行报错或停止暴露队列接口。这类条目会从目录移除,等后续检查恢复可用后再加入。
不拥有。每个模型页面仍由原作者托管和维护。你可以结合维护者、运行状态、关联模型和样例资产来判断是否适合你的工作流。
先看描述里的任务类型:图片转 3D、文字转 3D、网格重建、Gaussian splat、深度或查看器工作流。然后再看运行状态、作者、标签和关联模型。