Generate 3D assets from text
Text-to-3D and image-to-3D with PBR materials and engine-ready export for games, XR, and product visualisation.
3D stack.Prompt to engine-ready mesh
Tripo 3D v3.1
High-fidelity 3D asset generation with enhanced geometry and texture precision
Tripo 3D v3.1, production-ready assets from text or images with PBR materials and export to GLB, FBX, OBJ, and STL.
Tradeoff. $0.3 per text-to-3D generation, $0.4 for image-to-3D and multiview-to-3D.
Hunyuan 3D 3.1 Pro
High-fidelity 3D asset generation with multi-view reconstruction and PBR materials
TRELLIS.2
High-fidelity image-to-3D generative model with compact structured latents
SAM 3D Objects
Single-image 3D reconstruction for objects and scenes
Common questions
Generate — Tripo 3D v3.1; Build — Hunyuan 3D 3.1 Pro; Convert — TRELLIS.2; Reconstruct — SAM 3D Objects. Each row above expands with the model's live pricing, capability chips, and sample outputs.
TRELLIS.2 and SAM 3D Objects run on Runware's own optimized compute (the platform's open-weight tier, billed on compute time) — check each model page for weight availability and license terms before self-hosting. The remaining picks are partner-served.
Tradeoffs to consider
Tripo and Hunyuan generate stylised assets from prompts or concept images; TRELLIS.2 and SAM 3D Objects reconstruct what is in a photo. Pick by input: concept art goes to generation, real product photos go to reconstruction.
Hunyuan 3D 3.1 Pro starts at $0.225 for geometry; multiview input, PBR, and face-count control add $0.15 each, and normal-map generation is a separate $0.375 base task. A fully-optioned asset costs several times the base rate; decide which surfaces actually need PBR.
Production notes
Every price on this page is the model's published rate from the live Runware catalog, using the cheapest listed configuration unless stated otherwise. Prices vary with resolution, duration, quality tier, or token volume, so check the pricing table on each model page before estimating unit economics.
The Runware catalog does not publish per-model latency figures, so this page does not quote end-to-end timings. Where a model's own description commits to speed (for example sub-second generation or realtime streaming), that claim is repeated here. For anything else, benchmark the exact models in the Playground with your own payload sizes before committing to an SLA.
Models are addressed by versioned AIR identifiers, so a workflow pinned to specific model versions keeps producing the same behaviour as new versions ship. Adopt upgrades deliberately by re-running your evaluation set against the new version before switching production traffic.
About this collection
A blueprint bundles the small set of models you would actually wire together for one production use case, instead of ranking a whole category. Each pick covers one stage of the workflow and links to its model page for full schema, pricing, and examples.
Picks are the live models of the Best 3D collection (7 models), covering both directions the catalog supports: text-to-3D (5 live models ship the io:text-to-3d capability) and image-to-3D (7 ship io:image-to-3d). Export formats and material support come from each model's published description.