Arko-T: A Foundation Model for Text-to-Structured 3D Generation

arXiv 2026 59 method

TLDR

Arko-T is a 4B-parameter text-to-design model that generates editable parametric CAD programs from natural language, outperforming frontier LLMs at lower cost.

Reasoning

The paper presents a novel approach focusing on design-state alignment rather than just code executability, with strong benchmark results across 12 metrics. However, the abstract lacks details on real-world deployment or diverse part types, and the model's scalability to complex designs is not discussed.

Read-first score

Read-first score 59, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 83.

Recency 8%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Methodology quality 25%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=benchmark,metric,result

Topical relevance 42%

51.9

Uses existing LLM keyword relevance scores normalized to 0-100. AI for CAD,computer-aided design,neural CAD,generative CAD,parametric CAD,B-Rep,boundary representation,constructive solid geometry,CSG,sketch extrusion,CAD generation,CAD reconstruction,text-to-CAD,image-to-CAD,point cloud to CAD,CAD program

Reproducibility 25%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=artifact,code

Field roles

FrontierMethodology anchor

Rank sensitivity

Stability: volatile; rank range: 7.

Keyword Scores

parametric CAD

CAD generation

text-to-CAD

CAD program

computer-aided design

generative CAD

AI for CAD

neural CAD

sketch extrusion

CAD reconstruction

B-Rep

boundary representation

constructive solid geometry

CSG

image-to-CAD

point cloud to CAD

Deep Analysis

Innovations

Direct text-to-parametric-CAD-program generation, producing editable designs rather than static shapes
Alignment of data curation, code normalization, and execution-grounded supervision to a formal design state to preserve features, parameters, and construction logic
A 4B-parameter model that matches or exceeds frontier LLMs on structured CAD generation at one-tenth the cost

Methodology

Arko-T is a 4B-parameter model that maps natural language to executable parametric CAD programs. The pipeline aligns data curation, code normalization, and execution-grounded supervision with a formal notion of design state to ensure editability. It is evaluated against seven frontier LLMs across 12 metrics.

Key Results

Arko-T achieves the best score on 8 of 12 metrics and second-best on 3 more, while costing roughly one-tenth per benchmark compared to frontier LLMs.

Resources

Paper