Elf Bard
2026
This project followed a prompt-to-print pipeline: starting from a text prompt, generating a concept image with AI, and converting that image into a 3D model ready to refine in Blender. The idea was an elf bard seated on a tree trunk, playing a lyre, with the goal of capturing something graceful and dynamic in a single pose.
The first stage was prompting for form. A figure like this lives by its silhouette, so the concept had to read clearly from the outside in: the curve of the body, the angle of the lyre, and the way the pose suggests movement even at rest. Choosing the right concept also meant watching for the usual AI gotchas, where details look convincing in a flat image but fall apart once they need to exist as real geometry.
I used ChatGPT to generate the concept image and Tencent Hunyuan 3D to convert it into a mesh. From there the work moved into Blender for cleanup and a studio-clean shader and lighting setup, the kind of presentation that makes a piece feel considered rather than auto-generated.
Completed as part of Gesa Pickbrenner's course Prompt to Print: Use AI & Blender 3D to Create Your Dream Design on Skillshare.
