Overview
Layer supports a growing suite of 3D generation and refinement features through integrations with leading providers. This document outlines our supported capabilities, provider coverage, and how they fit into the wider Layer workflow.
Providers
Layer integrates with several 3D model providers, each with different strengths:
Tripo
Strengths: Stable mesh generation, consistent topology, supports parts generation and rigging.
Meshy
Strengths: Strong out-of-box visual quality; wide toolset including retexturing and remeshing to a clean quad mesh, available directly from the 3D viewer.
Rodin
Strengths: Strong visual and UV quality for single-model generation.
Hunyuan (Tencent)
Strengths: Unique features around segmenting and parts-based generation.
📘 Check the Base Models page, filtered to 3D, for the most current lineup, ratings, and pricing — the exact roster changes as new models are added.
Key Feature Areas
Image-to-3D: generate a 3D model from a source image, choosing a model based on speed vs. detail.
Face Limit control: set how many faces the generated mesh has — lower for lightweight, mobile-friendly models, higher for finer detail.
Retexturing: repaint a model's surface from a text prompt without changing its shape.
Remeshing / Quad Mesh: rebuild a model's topology into a clean quad mesh — the geometry DCC tools and game engines prefer for rigging, animation, and hand-editing.
Rigging & Animation: add a skeleton to a finished model and apply animation presets. Bipedal characters use a standard humanoid rig; non-bipedal models (like quadrupeds) are supported with a more generic bone structure.
Parts Generation: segment a model into editable sub-meshes, which you can view individually, hide, show, or explode apart for inspection.
The 3D Viewer
Every 3D model opens in a dedicated viewer with:
Orbit, pan, zoom, camera presets (Front, Top, Left, Right), and auto-rotate
Render modes — Textured, Solid, Normals, and Wireframe (for checking mesh topology)
Lighting, floor, and shadow toggles
A version ribbon along the bottom so you can compare the original model against retextured or remeshed revisions over time
Output Formats
3D models generate as .glb files — the standard format for browser preview and most provider outputs, saved to your Library like any other asset. If you need a different format for your engine or DCC pipeline, reach out to sales@layer.ai to confirm current export options.
What's next
The full generation walkthrough → Creating and refining 3D models
Choose the right model → Base Models