Releases: x-ili-x/stable-diffusion-webui-forge
WebUI Forge v1.0-fluxfix – Full support for Flux txt2img, img2img, inpainting, and Highres fix
🚀 Flux Fix Release (v1.0-fluxfix)
This release fully restores compatibility between Flux models and WebUI Forge, fixing everything that’s been broken since the SD3+ merge.
✅ Fixes included:
txt2img RuntimeError due to shape mismatch — solved
img2img producing garbage output — fixed
inpainting not working at all — resolved
Highres fix crashing or failing silently — now works
🔧 How:
Corrected model input shapes for Flux
Cleaned up logic across processing.py and related diffusion modules
Switched to a safer method for detecting inpainting/denoise capability
🧪 Tested and working with:
Custom SD1.5, SDXL and Flux checkpoints
MentatBot’s LoRA + ControlNet pipeline
High resolution batches up to 2048x2048
📌 Notes:
This fixes a 9-month-long bug that broke core features of Forge with Flux. If you’ve been holding off or dealing with band-aid solutions, this is the version to use.