3D Generation AI Models
Create 3D models from images or text, then retexture, segment and remesh them with Hunyuan3D, Tripo, Meshy, Rodin, Hi3D, TRELLIS and SAM 3.
All 3D Generation AI Models
Preview the output style, compare capabilities, then open any model for full specs, pricing, and a live Playground.
SAM 3 3D Align
SAM 3D Align by Meta places a SAM 3D Body mesh into the scene of its source image, optionally combining it with a SAM 3D Objects mesh into one GLB scene.
Hunyuan 3D 3.1 Part Splitter
Hunyuan 3D 3.1 Part by Tencent splits an FBX 3D model into its separate parts and returns one FBX file per part.
Tripo P1 Text to 3D
Tripo P1 is Tripo3D's text-to-3D model. It generates a 3D mesh with optional textures from a text description.
Hunyuan3D 2.0 Multi-View Turbo
Hunyuan3D 2.0 Multi-View Turbo by Tencent, the fast multi-view variant, generates a 3D mesh from front, back, and left view images, as a white mesh or with textures.
Hunyuan3D 2.0 Turbo
Hunyuan3D 2.0 Turbo by Tencent, the fast variant of Hunyuan3D 2.0, generates a 3D mesh from a single image, as a white mesh or with textures.
TRELLIS.2 Retexture
TRELLIS.2 Retexture by Microsoft paints new textures onto an existing 3D mesh (GLB, OBJ, PLY or STL) from a reference image.
Hunyuan3D 2.0 Mini
Hunyuan3D 2.0 Mini by Tencent, a smaller and cheaper version of Hunyuan3D 2.0, generates a 3D mesh from a single image, as a white mesh or with textures.
Hunyuan3D 2.0 Mini Turbo
Hunyuan3D 2.0 Mini Turbo by Tencent, the fast variant of Hunyuan3D 2.0 Mini, generates a 3D mesh from a single image, as a white mesh or with textures.
Hunyuan 3D 3.1 Smart Topology
Hunyuan 3D 3.1 Smart Topology by Tencent retopologizes a GLB or OBJ model into a clean triangle or quad mesh at a chosen polygon density.
Tripo P1 Image to 3D
Tripo P1 is Tripo3D's image-to-3D model. It reconstructs a 3D mesh with optional textures from a single image.
Hunyuan 3D 3.0 Text to 3D
Hunyuan 3D 3.0 by Tencent generates a textured, low-poly, or geometry-only 3D model from a text prompt.
Meshy Rigging
Meshy Rigging adds a humanoid skeleton to an existing character model and can also apply one animation preset.
Hunyuan 3D 3.1 Rapid Text to 3D
Hunyuan 3D 3.1 Rapid by Tencent quickly generates a textured or geometry-only 3D model from a text prompt.
Meshy 6 Text to 3D
Meshy 6 generates a 3D model from a text prompt, either as an untextured preview mesh or fully textured, with optional rigging and animation.
Hunyuan 3D 3.1 Rapid Image to 3D
Hunyuan 3D 3.1 Rapid by Tencent quickly turns a single image into a textured or geometry-only 3D model.
Hunyuan3D 2.0 Multi-View
Hunyuan3D 2.0 Multi-View by Tencent generates a 3D mesh from front, back, and left view images, as a white mesh or with textures.
SAM 3 3D Body
SAM 3D Body by Meta reconstructs the 3D shape and pose of the people in a single image and returns them as a GLB mesh.
Hunyuan3D 2.0
Hunyuan3D 2.0 by Tencent generates a 3D mesh from a single image, as a white mesh or with textures.
Meshy 6 Multi-Image to 3D
Meshy 6 reconstructs a 3D model from 1–4 views of the same object, with optional textures, PBR maps, auto-rigging and animation.
TRELLIS Multi-Image
TRELLIS by Microsoft generates a textured 3D mesh (GLB) from several views of the same object.
Rodin Gen-2 Image to 3D
Rodin Gen-2 by Hyper3D (Deemos) generates production-ready 3D assets from one to five reference images.
Tripo P2 Text to 3D
Tripo P2 is Tripo3D's latest flagship 3D generation model. It turns a text description into a production-ready mesh with optional PBR textures.
Tripo P2 Image to 3D
Tripo P2 is Tripo3D's latest flagship 3D generation model. It reconstructs a production-ready mesh with optional PBR textures from a single image.
Meshy 7.1 Text to 3D
Meshy 7.1 generates a textured, optionally rigged and animated 3D model from a text prompt.
Meshy 7.1 Multi-Image to 3D
Meshy 7.1 reconstructs a textured, optionally rigged 3D model from 1–4 views of the same object.
Meshy 7.1 Image to 3D
Meshy 7.1 turns a single image into a textured, optionally rigged and animated 3D model.
Rodin 2.5 Text to 3D
Rodin Gen-2.5 by Hyper3D (Deemos) generates game- and production-ready 3D assets from a text prompt.
Rodin 2.5 Image to 3D
Rodin Gen-2.5 by Hyper3D (Deemos) generates game- and production-ready 3D assets from one to five reference images.
Rodin 2.5 Fast Text to 3D
Rodin Gen-2.5 by Hyper3D (Deemos) generates game- and production-ready 3D assets from a text prompt, tuned for speed with the low tiers.
Rodin 2.5 Fast Image to 3D
Rodin Gen-2.5 by Hyper3D (Deemos) generates game- and production-ready 3D assets from one to five reference images, tuned for speed with the low tiers.
Tripo H3.1 Text to 3D
Tripo H3.1 is Tripo3D's text-to-3D AI model. Turn written descriptions into textured 3D objects with realistic geometry and materials.
Tripo H3.1 Multiview to 3D
Tripo H3.1 Multiview To 3d by Tripo3D - convert 2D images into 3D models with AI. Generate textured 3D assets from single photos for games, AR/VR, e-commerce, and digital content creation.
Tripo H3.1 Image to 3D
Tripo H3.1 by Tripo3D - convert 2D images into 3D models with AI. Generate textured 3D assets from single photos for games, AR/VR, e-commerce, and digital content creation.
Hunyuan 3D 3.1 Pro Text to 3D
Hunyuan 3d 3.1 Pro is Tencent's text-to-3D AI model. Turn written descriptions into textured 3D objects with realistic geometry and materials.
Meshy 6 Image to 3D
Meshy 6 turns a single image into a 3D model, with optional textures, PBR maps, humanoid auto-rigging and animation.
SAM 3 3D Objects
Sam 3 3d Objects by Meta - convert 2D images into 3D models with AI. Generate textured 3D assets from single photos for games, AR/VR, e-commerce, and digital content creation.
TRELLIS
TRELLIS by Microsoft generates a textured 3D mesh (GLB) from a single image using structured 3D latents.
Hunyuan 3D 3.1 Pro Image to 3D
Hunyuan 3d 3.1 Pro by Tencent - convert 2D images into 3D models with AI. Generate textured 3D assets from single photos for games, AR/VR, e-commerce, and digital content creation.
Hunyuan 3D 3.0 Image to 3D
Hunyuan 3d V3 is Tencent's image-to-3D AI model. Transform photographs into production-ready 3D meshes with accurate geometry and texture mapping.
TRELLIS.2
TRELLIS.2 by Microsoft generates a textured 3D model (GLB with PBR materials) from a single image.
Pixal3D
Pixal3D generates a high-fidelity textured 3D model (GLB) from a single image.
TripoSplat
TripoSplat by Tripo3D turns a single image into a 3D Gaussian splat for real-time rendering. Output is a .ply or .splat file.
Meshy Rigging Multi-Animation
Meshy Rigging Multi-Animation rigs an existing humanoid character model and generates up to 10 animation clips on the same rig.
Meshy 7 Multi-Image to 3D
Meshy 7 reconstructs a 3D model from 1–4 views of the same object, with optional textures, PBR maps, 2K geometry, auto-rigging and animation.
Meshy 7 Image to 3D
Meshy 7 turns a single image into a 3D model, with optional textures, PBR maps, 2K geometry, humanoid auto-rigging and animation.
Meshy 7 Text to 3D
Meshy 7 generates a 3D model from a text prompt, either as an untextured preview mesh or fully textured, with optional 2K geometry, rigging and animation.
Hitem3D Hi3D Image to 3D
Hi3D by Hitem3D generates a 3D model from a single image, with a selectable model version (v1.5, v2.0, v2.1, or portrait-tuned variants) and resolution tier.
Hitem3D Hi3D Multi-view to 3D
Hi3D by Hitem3D generates a 3D model from a front view plus optional back, left and right views, with a selectable model version and resolution tier.
Hitem3D Hi3D Multicolor
Hi3D Multicolor by Hitem3D converts a textured GLB model into a model with a limited set of flat colors, for multi-color 3D printing.
Hitem3D Hi3D Split
Hi3D Split by Hitem3D splits a 3D model into parts, either by a humanoid character template with optional joints or by granularity for general objects.
Hitem3D Hi3D Texture
Hi3D Texture by Hitem3D adds textures to an untextured GLB model using a reference image. It uses the v1.5 texture models.
Hitem3D Hi3D 3.0 Image to 3D
Hi3D 3.0 by Hitem3D reconstructs a detailed, watertight 3D model at 2048³ resolution from a single image, with optional texture and PBR materials.
Hitem3D Hi3D 3.0 Multi-view to 3D
Hi3D 3.0 by Hitem3D reconstructs a detailed 3D model at 2048³ resolution from a front view plus optional back, left and right views.
Meshy 6 Lite Retexture
Meshy 6 Lite Retexture is the lower-cost Meshy retexturing model: repaint an existing 3D model from a text or image style.
Meshy 6 Retexture
Meshy 6 Retexture repaints an existing 3D model from a text or image style, with optional PBR maps and up to 8K textures.
Meshy 7 Retexture
Meshy 7 Retexture repaints an existing 3D model from a text or image style, with optional PBR maps and up to 8K textures.
Meshy N1 Retexture
Meshy N1 Retexture repaints an existing 3D model from one reference image or up to four reference views, producing a 2K base color texture.
Meshy Relief
Meshy Relief turns a photo into a printable 3D relief (GLB and STL) with emboss or engrave carving and an optional curved bend.
Meshy T2.5 Image to 3D
Meshy T2.5 turns a single image into a clean Smart Topology triangle mesh (up to 30,000 triangles), with optional textures, rigging and animation.
Meshy T2.5 Text to 3D
Meshy T2.5 generates a clean Smart Topology triangle mesh (up to 30,000 triangles) from a text prompt, untextured in preview mode or textured in full mode.
Meshy T2 Image to 3D
Meshy T2 turns a single image into a clean Smart Topology triangle mesh (up to 15,000 triangles), with optional textures, rigging and animation.
Meshy T2 Text to 3D
Meshy T2 generates a clean Smart Topology triangle mesh (up to 15,000 triangles) from a text prompt, untextured in preview mode or textured in full mode.
MoGe-2
MoGe-2 by Microsoft estimates metric 3D geometry from a single image and returns a textured GLB mesh, a PLY point cloud, and depth, normal and mask maps.
MoGe-3
MoGe-3 by Microsoft estimates metric 3D geometry from a single image, refining depth with a sparse 3D refiner, and returns a textured GLB mesh, a PLY point cloud, and depth, normal and mask maps.
ReconViaGen 0.5
ReconViaGen 0.5 reconstructs a textured 3D model (GLB with PBR materials) from one or more views of the same object; more views give a more accurate result.
Idol
Idol by Sperid Labs turns a single image of an object into a 3D Gaussian splat (PLY) with a GLB point-cloud preview.
Tripo v2.5 Image to 3D
Tripo v2.5 by Tripo3D generates a textured 3D mesh from a single image, with optional HD textures, PBR materials and quad topology.
Tripo v2.5 Multiview to 3D
Tripo v2.5 by Tripo3D reconstructs a textured 3D mesh from a front view plus optional left, back and right views of the same object.
Tripo Remesh
Tripo Remesh rebuilds an existing 3D model with clean triangle or quad topology at a target face count. Accepts GLB, OBJ, FBX or STL input.
Tripo Segment
Tripo Segment splits an existing 3D model into named parts. Accepts GLB, OBJ, FBX or STL input.
Generate 3D assets with AI
3D generation models produce a textured mesh, usually as GLB, from a single image, several views, or a text prompt. Follow-up endpoints retexture, split into parts, remesh or rig the result.
Pick by input
- From an image: image-to-3d models such as Hunyuan3D 3.1, Tripo H3.1, Meshy 7, Rodin 2.5 and TRELLIS 2.
- From text: text-to-3d models including Hunyuan3D, Tripo and Meshy text endpoints.
- From an existing mesh: retexture (Meshy, TRELLIS 2), segment (Tripo, Hi3D split, SAM 3 align) and topology tools.
Start using 3D Generation AI Models on SandBase
One API key and one request shape for every model on this page.
- 01
Create an API key
Sign in to the SandBase Console and create a key on the API Keys page. Keep it in an environment variable and never commit it to source control.
TERMINALexport SANDBASE_API_KEY="YOUR_SANDBASE_API_KEY" - 02
Submit a run
Send a request to the Unified Run API with model tencent/hunyuan-3d/3.1/pro/image-to-3d. Only the model field changes between models on this page.
TERMINALcurl https://api.sandbase.ai/v1/run \ -H "Authorization: Bearer $SANDBASE_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model": "tencent/hunyuan-3d/3.1/pro/image-to-3d", "image": "https://static.sandbase.ai/examples/94e37e73-ddad-454e-99f6-6362fce34d94/input/64c93e7c-ae98-4ead-975e-c35b5e237d72.png"}' - 03
Fetch the result
The response includes a run id. Poll it every few seconds until the status is completed, then download the output URL.
TERMINALcurl https://api.sandbase.ai/v1/run/$RUN_ID \ -H "Authorization: Bearer $SANDBASE_API_KEY"
Questions, answered
What file format do 3D models return?
Most return a GLB file URL, often with a preview image. The model page lists the exact output fields.
Should I use one image or multiple views?
Multi-view endpoints, such as Tripo multiview, Meshy multi-image and Hi3D multi-view, usually reconstruct the back and sides more accurately than a single image.
Can I get separate parts of a model?
Yes. Hunyuan3D 3.1 Part, Tripo Segment and Hi3D Split break a mesh into parts.
How long does 3D generation take?
3D jobs run asynchronously and usually take longer than images. Submit to POST /v1/run and poll by id.
Build with 3D Generation AI Models.
One key, consistent APIs, and usage-based pricing across every model on this page.