Overview

Meshy-6 is a 3D generation model for creating meshes from text prompts or input images.

It is best suited to workflows that need stronger geometry quality, better hard-surface structure, more control over mesh topology and polygon budgets, and export-ready assets for games, prototyping, and 3D production pipelines.

Strengths

Cleaner Geometry

Meshy-6 is designed to produce cleaner and more usable geometry than earlier Meshy models. It is particularly strong for assets where structural consistency matters, helping reduce cleanup time after generation.

Better Hard-Surface Detail

The model is a strong fit for mechanical objects, props, and other hard-surface assets that benefit from sharper edges, clearer silhouettes, and more disciplined geometry.

Low-Poly Workflow Support

Meshy-6 includes a dedicated low-poly mode for generating lighter assets with cleaner polygon structure. This makes it useful for real-time workflows, game asset ideation, and pipelines where topology efficiency matters as much as visual detail.

Mesh Control

The API exposes meaningful control over topology and mesh density. Meshy-6 supports quad-dominant or triangle output, configurable polycount targets, adaptive decimation options, symmetry controls, and pose options such as A-pose and T-pose for suitable subjects.

Textured and PBR-Ready Output

Meshy-6 supports mesh-only and textured generation, with optional PBR materials. In supported workflows it can generate additional maps and cleaner texture outputs for custom lighting setups, making it more useful for downstream rendering and engine integration.

Capabilities

Text-to-3D

Meshy-6 supports text-to-3D generation through a preview and refine workflow. The preview stage generates the base mesh, and the refine stage textures that mesh for a more complete result.

Image-to-3D

The model supports image-to-3D generation from a reference image, with options for textured or untextured output depending on the workflow.

Multi-Image 3D Reconstruction

Meshy-6 also supports multi-image 3D generation, allowing several views of the same object to guide reconstruction for improved structure and consistency.

Production-Oriented Export

Meshy-6 supports common 3D export formats used in content pipelines, including GLB, FBX, OBJ, USDZ, and STL, with additional format support in some workflows.

Input and Output

• AIR ID: meshy:meshy@6
• Official API model ID: meshy-6
• Input: text prompts or one or more reference images depending on endpoint
• Output: 3D meshes with optional textures and PBR materials

Best Fit

• Game asset ideation and low-poly workflows
• Hard-surface props and product-style objects
• Character or creature base meshes that need cleaner structure
• Image-based 3D reconstruction from reference views
• 3D printing, prototyping, and engine-ready asset pipelines