p HA howto

3D Generation

AI 3D Models — Text and Image to Game-Ready Assets

3D generation is here. Not "coming soon." Promptha gives you access to 4 production-grade text-to-3D and image-to-3D models — Rodin, Hunyuan3D, Meshy, and Trellis. From concept to asset in under two minutes, at 15–25 credits per generation.

Available 3D Models

4 models, each with a distinct strength. Pick the one that matches your output requirements.

Rodin

Hyper3D ~15-20 credits

Game-ready 3D from text or image

Rodin generates production-quality 3D meshes from a single text prompt or reference image. Outputs are game-engine compatible with clean topology, ready for direct import into Unity, Unreal, or Blender.

Strengths

  • • Clean, game-ready mesh topology
  • • Accepts both text prompts and reference images
  • • Fast generation — typically under 60 seconds
  • • Supports GLB and OBJ export formats

Best For

Game asset creation Rapid prototyping Concept art to 3D conversion

Hunyuan3D v3

Tencent ~20-25 credits

PBR textures, professional quality

Tencent's Hunyuan3D v3 produces high-fidelity 3D models with physically-based rendering (PBR) textures out of the box. The output includes albedo, normal, roughness, and metallic maps — no manual texturing required.

Strengths

  • • Full PBR texture generation (albedo, normal, roughness, metallic)
  • • Professional-grade surface detail
  • • Consistent geometry with minimal artifacts
  • • Suitable for portfolio and production work

Best For

Product visualization Architectural rendering Professional portfolios

Meshy

Meshy AI ~20-25 credits

Full pipeline — text to 3D to rigging

Meshy covers the entire 3D pipeline: generate a model from text, apply textures, and rig it for animation. It is the most complete single-model solution for creators who need animated 3D assets without a team.

Strengths

  • • End-to-end pipeline: generation, texturing, and rigging
  • • Built-in animation rigging for characters
  • • Multiple art styles (realistic, stylized, low-poly)
  • • Direct export for game engines and 3D software

Best For

Character design and animation Indie game development Social media 3D content

Trellis

Microsoft ~15-20 credits

Versatile SLAT-based 3D generation

Microsoft's Trellis uses Structured Latent (SLAT) representations to generate diverse, high-quality 3D assets. It handles a wide range of object categories with strong geometric consistency and supports multiple output formats.

Strengths

  • • SLAT architecture for robust geometry
  • • Wide object category coverage
  • • Strong geometric consistency across views
  • • Flexible output format support

Best For

Research and experimentation Batch asset generation E-commerce product mockups

Use Cases

AI 3D generation fits into workflows across industries. Here is how different teams use it.

Game Developers

Game Development

Generate props, environment pieces, and characters from text descriptions. Skip weeks of manual modeling for indie and prototype projects.

Product Teams

Product Visualization

Turn product descriptions into 3D renders for e-commerce listings, pitch decks, and marketing materials. No photography studio required.

Architects & Designers

Architecture and Interiors

Generate furniture, fixtures, and decorative objects for architectural visualizations. Populate scenes without sourcing individual assets.

Character Artists

Character Design

Concept a character in text, generate the 3D model, rig it for animation. Iterate on designs in minutes instead of days.

Product Designers

Rapid Prototyping

Visualize physical products in 3D before manufacturing. Test form factors, proportions, and aesthetics without CAD expertise.

Educators

Education and Training

Create 3D models of anatomical structures, mechanical parts, or historical artifacts for interactive learning materials.

Getting Started

Four steps from concept to exported 3D asset.

01

Choose a Model

Pick the 3D model that fits your use case. Rodin for game assets, Hunyuan3D for PBR textures, Meshy for rigging, Trellis for versatility.

02

Describe or Upload

Write a text prompt describing the object you want, or upload a reference image. Be specific about style, material, and proportions.

03

Generate

Hit generate. Most models return a complete 3D asset in 30 to 90 seconds. Preview the result in the built-in 3D viewer.

04

Export

Download your model in GLB, OBJ, or other supported formats. Import directly into Unity, Unreal, Blender, or your 3D tool of choice.

Tips for Better Results

Practical advice for getting the most out of AI 3D generation.

Be Specific About Materials

Instead of "a sword," try "a medieval broadsword with a worn leather grip and tarnished steel blade." Material descriptions directly influence texture quality.

Use Reference Images When Possible

Image-to-3D consistently produces more accurate results than text-to-3D for complex or specific objects. A single reference image is worth a paragraph of description.

Match Model to Output Needs

Need PBR textures? Use Hunyuan3D. Need animation-ready rigging? Use Meshy. Need fast game assets? Use Rodin. Each model has a distinct strength.

Iterate on Prompts

If the first result is close but not right, refine the prompt rather than starting over. Add detail about the part that needs improvement — shape, texture, scale, or style.

Pricing

Cost per Generation

15–25 credits

Exact cost depends on the model and output complexity. Rodin and Trellis are on the lower end (15–20 credits). Hunyuan3D and Meshy, with their richer outputs, are on the higher end (20–25 credits).

Availability

3D generation is available on the Creator tier ($12/month) and above. All 4 models are included with no additional per-model fees.

View pricing tiers →

Start Generating 3D Assets

11 3D models. From concept to asset. Open Promptha, describe what you need, and export a game-ready 3D model in under two minutes.

Open Promptha