Learning Timeline
Key Insights
Storage Maintenance Tip
Ensure you have at least 20GB of free space before starting the download, as the Flux model (11GB) and Text Encoders (5GB) have relatively large file sizes.
Troubleshooting: Model Not Appearing
If the model does not appear in the dropdown list after downloading, make sure to click the 'Refresh' button in the menu or press the 'R' shortcut. If it still doesn't show up, double-check that the file is placed in the correct sub-folder within the 'models' directory.
Step by Step
How to Set Up Flux Models and Encoders in ComfyUI
- Download the 'Flux Create Dev FP8 Scaled' model file (approx. 11GB).
- Open your ComfyUI installation folder.
- Place the model file into the directory: ComfyUI > models > diffusion_models.
- Download the 'Clip L' Text Encoder file (246MB).
- Place the file into the directory: ComfyUI > models > clip.
- Download one of the additional Text Encoders (FP16 or FP8) depending on your RAM/VRAM capacity.
- Save the file into the same directory: ComfyUI > models > clip.
- Download the 'VAE' file (327MB).
- Place the VAE file into the directory: ComfyUI > models > vae.
- Open the ComfyUI interface.
- Press the 'R' key on your keyboard to run the 'Refresh' function so the newly downloaded models appear in the dropdown list.
- Select the model you just added from the dropdown menu on the relevant node in the workflow.
