# How viable is LLM-assisted 3D modeling in Blender? > Source: > Published: 2026-09-03 04:19:38+00:00 GLM-5.3-Flash is an excellent model. Running it locally on 4 x RTX 6K Pros with great results. Was using GLM-5.2 locally before and I can see the difference. It seems faster and smarter. The vision capabilities are also extremely helpful for my workflows.