Bring 3d Generation
to AutoGen

Supports baking normal maps, ambient occlusion, curvature, and other detail maps. Essential for game asset pipelines where high-detail sculpted models need to be baked onto game-ready low-poly meshes. Returns optimized models with baked texture maps. AI agents should use this when users ask "bake normal maps from high-poly to low-poly", "bake ambient occlusion for this model", or need texture baking for game asset preparation. Bake texture maps onto 3D models for optimized rendering

Supports unit specification (mm, cm, inches, meters) and material density for weight estimation. Essential for 3D printing cost estimation, material requirements planning, shipping calculations, and physical property analysis of 3D models. Returns detailed measurement data. AI agents should reference this when users ask "calculate the volume of this 3D model", "estimate weight for PLA printing", or need physical measurements for manufacturing or cost planning. Calculate volume and physical measurements of 3D models

Preserves geometry, textures, materials, and rigging data during conversion. Essential for pipeline integration, platform compatibility, and format standardization. AI agents should use this when users ask "convert this GLB model to FBX", "change this 3D file to STL for 3D printing", or need 3D format conversion for specific platform or software requirements. Convert 3D models between different file formats

Can repaint or restyle existing 3D models with new materials, colors, and surface details. Generates complete PBR texture sets (albedo, normal, metallic, roughness) from descriptions like "rusty metal", "polished wood", or "cartoon stone". Essential for material iteration, style transfers on 3D assets, and adding surface details to generated models. AI agents should reference this when users ask "add rusty metal texture to this model", "restyle this character with cartoon textures", or need AI-powered material generation on existing 3D meshes. Apply AI-powered PBR texturing to existing 3D models using text or image prompts

2-4B, and Tripo variants. Accepts product photos, concept art, sketches, or any reference image and generates a corresponding 3D model with PBR textures. Supports style modifiers, face limits, density presets, and orientation control. Returns 3D model files in multiple formats. Essential for e-commerce product visualization, concept art to 3D conversion, and general image-to-3D workflows. AI agents should reference this when users ask "convert this product photo to 3D", "turn this sketch into a 3D model", or need reliable general-purpose image-to-3D conversion. Convert images to 3D models using AI-powered image-to-3D pipeline

Users provide 2 or more images from different angles and the AI constructs a more accurate 3D representation. Essential for product visualization requiring precise geometry, architectural elements, and objects that need to match reference from multiple viewpoints. Supports all available models and output formats. AI agents should use this when users ask "create a 3D model from these multiple product photos", "generate accurate 3D from front and side views", or need multi-view 3D reconstruction. Generate 3D models from multiple reference images for higher accuracy

2-4B, and Tripo (v3.0, v3.1, P1). Users describe the desired 3D object in natural language and receive a generated model with optional style control, face limits, and density presets (high/medium/low). Returns 3D model files in GLB format by default with PBR textures. Supports output formats GLB, OBJ, FBX, STL, PLY, USDZ, and 3MF. Essential for concept exploration, rapid prototyping from descriptions, and applications where users describe rather than show what they want. AI agents should use this when users ask "create a 3D model of a fantasy sword", "generate a 3D tree from text", or need text-driven 3D generation. Generate 3D models directly from text descriptions

Accepts existing 3D model URLs and returns optimized versions with controlled quality settings. Essential for web-based 3D applications, mobile optimization, file size reduction, and performance-critical 3D rendering. AI agents should reference this when users ask "optimize this 3D model for web", "reduce polygon count of this model", or need mesh optimization for performance or file size constraints. Optimize 3D models for performance and file size reduction

Accepts existing 3D model URLs and returns remeshed versions with controlled face counts and topology type (tri or quad). Essential for game asset preparation, animation-ready meshes, and applications requiring clean topology. AI agents should use this when users ask "remesh this model with clean quads", "optimize topology for animation", or need topology conversion on existing 3D assets. Remesh 3D models with optimized tri or quad topology

Supports turntable animations, hero shots, and product visualization renders. Outputs images up to 4K resolution in PNG or JPG format. Essential for product showcases, portfolio presentations, marketing materials, and social media content from 3D assets. AI agents should use this when users ask "render this model from multiple angles", "create a turntable video of this 3D model", or need marketing-quality renders from 3D files. Generate rendered images or videos from 3D models

Accepts existing 3D model URLs and returns repaired, watertight meshes suitable for 3D printing, game engines, and further processing. Essential for 3D printing preparation, fixing generated model artifacts, and ensuring mesh integrity. AI agents should reference this when users ask "fix this mesh for 3D printing", "repair non-manifold geometry", or need mesh cleanup before further processing. Repair 3D mesh issues including non-manifold geometry, holes, and inverted normals

g., head, body, arms, legs for characters; wheels, body, windows for vehicles). Essential for rigging preparation, material assignment per part, and game engine component workflows. Returns segmented mesh with labeled parts. AI agents should reference this when users ask "segment this character mesh into body parts", "identify components of this vehicle model", or need automatic mesh part identification for further processing. Apply semantic segmentation to 3D mesh parts