Image or text to textured 3D model
3D Generator
The 3D Generator creates textured 3D models from a reference image, multi-view photos, a batch of images, or a text description. Three AI model providers are available-each with different strengths in geometry quality, texture fidelity, and export options. Generated models can be downloaded as GLB, FBX, OBJ, or STL and imported directly into Blender, Unreal Engine, Unity, Spline, or any 3D-capable tool.
01
Choose a model and input mode
Select your AI model (Hunyuan 3D, Tripo, or Meshy) and switch to the input mode that matches your source-Image, Multi-view, Batch, or Text.
02
Upload or describe
Drag in your source image(s) or type a text prompt. Adjust advanced settings-texture quality, face count, topology-if needed.
03
Generate
Click Generate. Generation typically takes 30–120 seconds depending on the model and settings. Progress is shown in real time.
04
Review in the viewer
The completed model loads in the interactive viewer. Toggle surface modes, wireframe, lighting, and background. Check face and vertex counts in the mesh stats panel.
05
Post-process (optional)
Run Retopology to reduce polygon count for real-time use, or Retexture to replace the generated texture with a new one driven by a prompt or image.
06
Export
Choose an export format (GLB, FBX, OBJ, STL) and download. All formats include baked textures except STL.
Credit cost: Base generation costs vary by model. Retopology and Retexture passes each consume additional credits. If generation fails, the credit is automatically refunded.