Image-to-3D and Text-to-3D Asset Generation
Generating textured 3D meshes from one or a few images or a text prompt.
Tier Grades
Tier A: Frontier AI labs
feasible
A 74-author team built Hunyuan3D 2.0, frontier-lab scale.
Tier B: Indonesian AI labs
not feasible
Pretraining is not feasible for an Indonesian lab.
Tier C: Institution units such as ITB
stretch
A university unit can fine-tune small open models such as TripoSR.
Tier D: Organisations such as SEMAI (PoT, PoC, PoV)
stretch
Product 3D views and digital twins are possible pilots, but output quality needs checking.
Where this comes from
Tiers: a) frontier AI labs, b) Indonesian AI labs, c) institution units such as ITB, d) organisations such as SEMAI delivering proof of technology, concept and value. Each rationale cites the anchors below.
Anchors
- ComputeTraining compute is not in the sources read.
- DataNot in the sources read; Hunyuan3D 2.0 has a shape model (Hunyuan3D-DiT) and a texture model (Hunyuan3D-Paint).
- Team size (proxy)Proxy only (author count of an exemplar paper, not a headcount): 74 authors on the Hunyuan3D 2.0 report (arXiv 2501.12202).
- Time horizonMonths.
- Shelf-life signalsProduction-ready PBR meshes from one image, Generated 3D scenes rather than single objects
- Pipeline roleTier A releases 3D generators; Tier C tests them; Tier D uses them for product and asset previews.
Where this comes from
Team size is a proxy (author counts of exemplar papers), not a headcount.
Who and What
Exemplars
Observed actors
Linked models
Hunyuan3D 2.03Dexemplar
TRELLIS (image-large)3DexemplarInstantMesh3Dexemplar
TripoSR3Dexemplar
Lyra-2.03Dproduced byPixal3D3Dproduced byTRELLIS.2-4B3Dproduced by
Meshy 6 Preview3Dproduced byHunyuan3D-Omni3Dproduced by
Tripo 3.03Dproduced byHunyuan3D-2.13Dproduced byHunyuan3D 2mini3Dproduced byHunyuan3D 2mv3Dproduced byTripoSG3Dproduced by
LLaMA-Mesh3Dproduced byHunyuan3D 1.03Dproduced bySV3D3Dproduced byStable Zero1233Dproduced by
Shap-E3Dproduced byEG3D3Dproduced by
Mesh R-CNN3Dproduced by
Rodin Gen-23Dproduced by
Genie3Dproduced bySAM 3D Objects3Dproduced byLinks
R&D TopicImage-to-3D and Text-to-3D Asset Generation
ModelsTripoSRStability AI, 3DTRELLIS (image-large)Microsoft AI, 3DHunyuan3D 2.0Tencent, 3DInstantMeshTencent, 3D+23 more
TripoSRStability AI, 3DTRELLIS (image-large)Microsoft AI, 3DHunyuan3D 2.0Tencent, 3DInstantMeshTencent, 3D+23 moreUse Casesnone recorded
Who Runs Itnone recorded
Models 27
Hunyuan3DTencent, 3Dproduced byEG3DNVIDIA, 3Dproduced byHunyuan3D 1.0Tencent, 3Dproduced byHunyuan3D 2.0Tencent, 3DexemplarHunyuan3D 2miniTencent, 3Dproduced byHunyuan3D 2mvTencent, 3Dproduced byHunyuan3D-OmniTencent, 3Dproduced byLLaMA-MeshTsinghua University, NVIDIA, 3Dproduced bySAM 3D ObjectsMeta, 3Dproduced bySV3DStability AI, 3Dproduced byTRELLIS (image-large)Microsoft AI, 3DexemplarTRELLIS.2-4BMicrosoft AI, 3Dproduced by+15
TripoSRStability AI, VAST, 3DexemplarGenieLuma AI, 3Dproduced byHunyuan3D-2.1Tencent, 3Dproduced byInstantMeshTencent, 3DexemplarLyra-2.0NVIDIA, 3Dproduced byMesh R-CNNMeta, 3Dproduced byMeshy 6 PreviewMeshy LLC, 3Dproduced byPixal3DTencent, 3Dproduced byRodin Gen-2Deemos Technology, 3Dproduced byShap-EOpenAI, 3Dproduced bySPAR3DStability AI, 3Dproduced byStable Fast 3DStability AI, 3Dproduced byStable Zero123Stability AI, 3Dproduced byTripo 3.0VAST, 3Dproduced byTripoSGVAST, 3Dproduced byPapers
- DreamFusion: Text-to-3D using 2D DiffusionBen Poole et al., arXiv preprint, 2022 founding
- Hunyuan3D 2.0: Scaling Diffusion Models for High Resolution Textured 3D Assets GenerationZibo Zhao et al., arXiv preprint, 2025 state of the art
- Structured 3D Latents for Scalable and Versatile 3D GenerationJianfeng Xiang et al., arXiv preprint, 2024 state of the art
- Zero-1-to-3: Zero-shot One Image to 3D ObjectRuoshi Liu et al., arXiv preprint, 2023 state of the art
- LRM: Large Reconstruction Model for Single Image to 3DYicong Hong et al., ICLR 2024 (per arXiv comment), 2023 state of the art
Where this comes from
The papers to read first: where the direction began, its current best, and a survey when one exists.
Sources
- arxiv.org/abs/2501.12202 https://arxiv.org/abs/2501.12202
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Image-to-3D and Text-to-3D Asset Generation", https://aurion.semai.id/topics/image-to-3d-asset-generation/, records of October 2026.