Qwen Image 2.1 gets ComfyUI prompt-rewriting nodes
A post describes a ComfyUI add-on that expands short prompts into detailed English prompts for Qwen Image 2.1, supporting both image generation and editing.
TLDR
A post sharing the ComfyUI project says it packages official PE-T2I and PE-I2I prompt enhancement into nodes—not a new image model. Separately, an early Qwen Image 2.1 tester reports creating three images, including a 2048×1152 image with text and a 1024×1024 transparent PNG. The tester says Japanese text rendered without garbling and that a transparent PNG could be created through prompting alone.
Combined views
—
2 Sources, first seen ago
