Hyper3d Rodin V2.5 Image To 3d

Hyper3d Rodin V2.5 Image To 3d

Playground

Hyper3D Rodin v2.5 is a fast AI image-to-3D generation model that turns reference images into production-ready 3D assets with clean meshes, UVs, and textures. Ready-to-use REST inference API for game assets, e-commerce 3D models, product visualization, AR/VR content, digital twins, creative design, and professional 3D asset generation workflows with simple integration, no coldstarts, and affordable pricing.

Features

Hyper3D Rodin V2.5 Image-to-3D

Hyper3D Rodin V2.5 Image-to-3D generates 3D assets from one or more reference images, with optional text guidance and flexible controls for mesh density, texture output, geometry behavior, symmetry, pose, and export format. It is suitable for product concepts, stylized objects, character prototyping, game assets, and other image-to-3D workflows where you want direct control over geometry and material output.

Why Choose This?

• Image-guided 3D generation
  Turn one or more reference images into a 3D asset with minimal setup.

• Flexible mesh and quality options
  Choose between quad and triangle outputs across multiple density levels.

• Material and texture control
  Generate PBR, shaded, full material sets, or geometry-only output depending on your workflow.

• Geometry behavior settings
  Use geometry_instruct_mode, symmetry options, and micro-detail settings to shape how the asset is built.

• Multiple export formats
  Export as glb, usdz, fbx, obj, or stl.

• Production-ready workflow
  Useful for concept art, product prototyping, game asset generation, and 3D content pipelines.

Parameters

How to Use

1. Upload your reference images — provide 1–5 images of the object, character, or asset you want to generate.
2. Add a prompt (optional) — describe the asset if you want to guide shape, style, or design intent more explicitly.
3. Choose a tier — use the default tier for general work, or select Gen-2.5-Extreme-High when you need the highest-quality generation.
4. Choose mesh and format settings — set quality_and_mesh, geometry_file_format, and material based on your downstream use.
5. Adjust geometry and texture controls (optional) — fine-tune texture_mode, geometry_instruct_mode, is_symmetric, hd_texture, and related settings.
6. Add HighPack (optional) — enable addons = HighPack if you need higher-end texture and geometry output.
7. Submit — run the model and download the generated 3D asset.

Example Use Case

Generate a clean 3D product model from multiple reference photos, with PBR materials and quad mesh output for downstream rendering or e-commerce visualization.

Pricing

Pricing is charged per generation.

Billing Rules

• Base price is $0.40 per generation
• Gen-2.5-Extreme-High costs 2× the base price
• HighPack adds an additional $0.80 per generation
• HighPack and Gen-2.5-Extreme-High stack together
• Other parameters do not affect pricing

Best Use Cases

• Product concept modeling — Build 3D product shapes from reference photos.
• Game asset prototyping — Generate early 3D assets for props, items, and stylized objects.
• Character base generation — Create starting meshes for stylized or realistic character workflows.
• AR/VR content preparation — Build assets for immersive applications and previews.
• Creative 3D ideation — Rapidly test multiple visual directions before manual refinement.

Pro Tips

• Upload clean, well-lit reference images for better shape and texture consistency.
• Use multiple views when structure accuracy matters.
• Use faithful geometry mode when image fidelity matters more than creative interpretation.
• Use creative when you want the model to take more design liberties.
• Choose quad output when downstream retopology or sculpting workflows matter.
• Use texture_delight if you want cleaner textures with less baked lighting.
• Only enable is_micro when using Gen-2.5-Extreme-High, since it has no effect otherwise.
• Add HighPack only when you need the extra texture and geometry quality, since it increases cost significantly.

Notes

• images is required.
• You can upload up to 5 reference images.
• prompt is optional; if omitted, the model can infer guidance from the images.
• Gen-2.5-Extreme-High doubles the base generation cost.
• HighPack adds an extra fixed charge on top of the selected tier.
• is_micro is only effective with Gen-2.5-Extreme-High.
• geometry_file_format defaults to glb.
• material defaults to All.
• quality_and_mesh defaults to 500K Triangle.

Related Models

• Hyper3D Rodin V2.5 Text-to-3D — Generate 3D assets directly from text prompts.
• Other Hyper3D Rodin workflows — Useful when you need different 3D generation paths or input modalities.
• 3D asset generation pipelines — Useful when image-to-3D is only one part of a broader modeling workflow.

# Submit the task
curl --location --request POST "https://api.wavespeed.ai/api/v3/hyper3d/rodin-v2.5/image-to-3d" \
--header "Content-Type: application/json" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}" \
--data-raw '{
    "tier": "Gen-2.5-Medium",
    "use_original_alpha": false,
    "geometry_file_format": "glb",
    "material": "All",
    "quality_and_mesh": "500K Triangle",
    "geometry_instruct_mode": "faithful",
    "hd_texture": false,
    "texture_delight": false,
    "is_symmetric": "unknown",
    "is_micro": false,
    "ta_pose": false,
    "preview_render": false
}'

# Get the result
curl --location --request GET "https://api.wavespeed.ai/api/v3/predictions/${requestId}/result" \
--header "Authorization: Bearer ${WAVESPEED_API_KEY}"