Skip to content
note · kongo_jun · 图像视频创作分享·· 2026-10-20AI score55

Testing Qwen-Image-2.1's Multiple-Angles LoRA in ComfyUI

ComfyUIでQwen-Image-2.1のMultiple-Angles-LoRAを試してみた

AI briefing

The author set up and tested Qwen-Image-2.1's Multiple-Angles LoRA in ComfyUI, which can change the angle of the photographed subject via prompts. In testing, a single image took 23 seconds to generate and used 21GB of VRAM, but this LoRA currently focuses only on subject angle changes, the background disappears, and the output is fairly random. The article provides the custom node installation command, the placement paths for the model and LoRA files, and the workflow file.

Open original note.com