Hey, have you heard about the latest update to the WAN 2.2 Text2Image Custom Workflow? It’s a game-changer! After days of hard work, the new version is packed with exciting features that take image generation to the next level.
The creator has added a base model selector, individual activators for SageAttention and Torch Compile, and an ‘Image Style Loras’ panel to change the style of the generated image. You can even select the total steps, with automatic calculation for each KSampler. That’s not all – there’s also a ‘Prompt variation helper’ option to get more variation in the result, and the ability to use VAE or Tiled VAE.
The generated image is now upscaled to x2 by default, and there are new settings in KSamplers to prevent image defects. Plus, you can choose from new image enhancement options, including Instagram filters. And if you want to take it to the next level, you can upscale to x2 or x8 (up to 30k resolution).
The possibilities are endless with this new update. The creator is already thinking about what’s next – maybe using Qwen as the initial step or adding image2image control. For now, take some time to explore the new features and see what amazing creations you can come up with!