muapi Logomuapi Logo
muapi Logomuapi Logo

Meshy 6 Multi-Image to 3D: AI 3D Model API

Reconstruct a 3D mesh from 1-4 reference images. Multi-view inputs produce more accurate geometry for complex or asymmetric objects. Reconstruct a 3D mesh from 1-4 reference images. Multi-view inputs produce more accurate geometry for complex or asymmetric objects, while still supporting single-image submissions.

Interactive model controls

Reconstruct a 3D mesh from 1-4 reference images. Multi-view inputs produce more accurate geometry for complex or asymmetric objects.

📝

Overview

About this model

Reconstruct a 3D mesh from 1-4 reference images. Multi-view inputs produce more accurate geometry for complex or asymmetric objects, while still supporting single-image submissions.

1Asymmetric Objects: Capture front / side / back angles for accurate reconstruction.
2Character Design: Build base meshes from a model sheet with multiple views.
3Heritage / Artefacts: Reconstruct objects from a handful of photographs.
💰

Pricing & Value

Cost analysis

muapiapp$0.50 per generation

Flat rate regardless of polygon count or topology.

Fal.ai$1.20 per generation

Approximately 58% cheaper on muapiapp.

ReplicateNot available

This model is not currently available on Replicate.

** Competitor pricing is estimated based on similar model architectures and usage tiers.

⚙️

Technical Details

Configuration schema

Image URLsarray

1-4 reference images from different angles.

Default Valuehttps://cdn.muapi.ai/assets/meshy-6-multi-image-to-3d-1.jpg
Generate Textureboolean

Generate textures alongside the mesh.

Default Valuetrue
TopologyEnum (2 options)

Output mesh topology.

Default Valuetriangle
Target Polygon Countint

Target polygon count for the output mesh.

Default Value30000
Remeshboolean

Enable remesh phase.

Default Valuetrue
SymmetryEnum (3 options)

Symmetry detection mode.

Default Valueauto
Enable PBRboolean

Generate PBR maps.

Default Valuefalse
Pose ModeEnum (3 options)

Force humanoid characters into a standard pose.

Default Value
Texture Promptstring

Optional text to guide the texturing pass.

Default Valueweathered metal with rust accents
📖

Implementation Guide

Developer documentation

How to Use Meshy 6 Multi Image to 3D

  1. Upload 1-4 reference images: More views give better geometry; aim for 3-4 angles when possible.

  2. Choose whether to texture: should_texture: true (default) retains image colours.

  3. Pick topology and polygon count: triangle or quad; polycount from 100 to 300,000.

  4. Optional tuning: pose_mode for humanoid pose normalisation, enable_pbr for PBR materials, or texture_prompt to guide texturing.

  5. Submit: Poll until the prediction returns a .glb URL.

❓

Common Questions

Frequently asked

How many images should I provide?

3-4 images from different angles produce the most accurate reconstruction. 1-2 images still work but with less geometric fidelity.

Do my images need to be the same lighting and background?

Yes - consistent lighting and clean, similar backgrounds across views produce the cleanest results.

Is this better than the single-image variant?

For asymmetric or complex objects, yes. For simple front-facing items, single-image is usually sufficient and cheaper to upload.