Hunyuan3D V2 Multi-View

Model Information

Display Name: Hunyuan3D V2 Multi-View

API Model ID: tencent/hunyuan3d-v2-multi-view

Category: Image To 3D

Description: Tencent Hunyuan3D V2 Multi-View reconstructs accurate 3D models from three input views using a decoupled dual-model architecture. Combines a 2.6B parameter geometry model (Hunyuan3D-DiT) with a 1.3B parameter texture synthesis model (Hunyuan3D-Paint) for high-fidelity results in approximately 30 seconds. **Key Features:** - Three-view input: front, back, and left images - Decoupled geometry + texture architecture - 2.6B parameter geometry model (Hunyuan3D-DiT) - 1.3B parameter texture model (Hunyuan3D-Paint) - High-resolution 4K textured output - ~30 second generation time - GLB format output **Capabilities:** - Multi-view image-to-3D reconstruction - High-fidelity geometry reconstruction - 4K texture synthesis - Accurate shape and detail preservation - Consistent texture mapping across views **Best For:** - Quick 3D reconstruction from photos - Product digitization and visualization - E-commerce 3D product models - AR/VR asset creation - 3D scanning alternative from photos - Rapid 3D prototyping **Input Requirements:** - Front view image (required) - Back view image (required) - Left view image (required) - All three views must be provided **Technical Specs:** - Geometry Model: Hunyuan3D-DiT (2.6B parameters) - Texture Model: Hunyuan3D-Paint (1.3B parameters) - Architecture: Decoupled dual-model pipeline - Texture Resolution: Up to 4K - Generation Time: ~30 seconds - Output Format: GLB **Pricing:** - Flat rate per generation - No additional parameter-based costs **Limitations:** - Exactly 3 images required (front, back, left) - No text prompt support - Best results with consistent lighting across views - No configurable quality or geometry options

How to Use This Model

To use Hunyuan3D V2 Multi-View via the HInow.ai API, use the model ID: tencent/hunyuan3d-v2-multi-view

API Request Example (Image Generation)


POST https://api.hinow.ai/v1/images
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

{
  "model": "tencent/hunyuan3d-v2-multi-view",
  "prompt": "Your image description here"
}
              

Pricing

  • total: $0.028

Quick Reference

To use this model, set: "model": "tencent/hunyuan3d-v2-multi-view"

Featured: Yes

Documentation: https://hinow.ai/models/tencent/hunyuan3d-v2-multi-view

API Endpoint: https://api.hinow.ai/v1

Back to Models

Hunyuan3D V2 Multi-View

Featured

tencent/hunyuan3d-v2-multi-view

$0.028
per image

About

Tencent Hunyuan3D V2 Multi-View reconstructs accurate 3D models from three input views using a decoupled dual-model architecture. Combines a 2.6B parameter geometry model (Hunyuan3D-DiT) with a 1.3B parameter texture synthesis model (Hunyuan3D-Paint) for high-fidelity results in approximately 30 seconds.

Key Features:

  • Three-view input: front, back, and left images
  • Decoupled geometry + texture architecture
  • 2.6B parameter geometry model (Hunyuan3D-DiT)
  • 1.3B parameter texture model (Hunyuan3D-Paint)
  • High-resolution 4K textured output
  • ~30 second generation time
  • GLB format output

Capabilities:

  • Multi-view image-to-3D reconstruction
  • High-fidelity geometry reconstruction
  • 4K texture synthesis
  • Accurate shape and detail preservation
  • Consistent texture mapping across views

Best For:

  • Quick 3D reconstruction from photos
  • Product digitization and visualization
  • E-commerce 3D product models
  • AR/VR asset creation
  • 3D scanning alternative from photos
  • Rapid 3D prototyping

Input Requirements:

  • Front view image (required)
  • Back view image (required)
  • Left view image (required)
  • All three views must be provided

Technical Specs:

  • Geometry Model: Hunyuan3D-DiT (2.6B parameters)
  • Texture Model: Hunyuan3D-Paint (1.3B parameters)
  • Architecture: Decoupled dual-model pipeline
  • Texture Resolution: Up to 4K
  • Generation Time: ~30 seconds
  • Output Format: GLB

Pricing:

  • Flat rate per generation
  • No additional parameter-based costs

Limitations:

  • Exactly 3 images required (front, back, left)
  • No text prompt support
  • Best results with consistent lighting across views
  • No configurable quality or geometry options

Capabilities

Image To 3D

Code Examples

curl -X POST https://api.hinow.ai/v1/images \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $HINOW_API_KEY" \
  -d '{
    "model": "tencent/hunyuan3d-v2-multi-view",
    "prompt": "Transform this image into cyberpunk style",
    "images": [
      "https://example.com/image1.jpg",
      "data:image/png;base64,iVBORw0KGgo..."
    ]
  }'