Updated: Jan 28, 2026
styleDownload
1 variant available
The FLUX.1 [dev] Model is licensed by Black Forest Labs. Inc. under the FLUX.1 [dev] Non-Commercial License. Copyright Black Forest Labs. Inc.
IN NO EVENT SHALL BLACK FOREST LABS, INC. BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.
Workflow for weak PCs. For Flux 2 Klein, Qwen Image of different versions and other models.
Step 1: Make a backup of your installed libraries.
Update pip. Update libraries. Build the entire base around PyTorch (2.10.0+cu128). Update the library torchvision and all others that require updates. Update the xFormers library to xFormers 0.0.34 for stability and rendering speed. The CUDA version must be CUDA 12.8 or higher.
Step 2: Clear the cache before running ComfyUI.
Run cmd or powershell # Go to the ComfyUI root # 1. Delete ALL pycache folders and .pyc files in ALL custom nodes # 2. Delete the ComfyUI Manager cache (it often stores service information) # 3. (Optional) You can clear the ComfyUI interface cache. The neural network will write the code for you, save it.
Step 3 If you have problems with MultiGPU, there is a solution in the archive. Update or install the workflow nodes. If you have a portable version of ComfyUI, update it from the update folder.