
Currently, 'text to 3D mesh' models are emerging like Shap-E AI from OpenAI. They could maybe provide material for game engines but lack the features and accuracy to be used in industrial grade mecanical projets.
Will we see a 'text to 3D CAD' or 'image to 3D CAD' system powered by AI able to translate any reasonably complex part description into industrial grade CAD models before 2027 ?
The system would need to feature at least:
- Part modeled with correct geometry (size, shapes), and parametric features (holes, fillets, chamfers, bores, etc...) according to the input description. 
- Usable in at least one CAD software 
- Part must be ready for CAD assembly 
@ConnorTabarrok Very promising, if this product release and is able to generate a good range of part with various geometries and correct features this could be it !