AlphaVAE replaces the standard FLUX VAE with an RGBA-capable VAE, enabling direct transparent image generation without post-processing background removal.

Nodes

Node	Description
`AlphaVAE Loader`	Load AlphaVAE model (diffusers format)
`AlphaVAE Decode`	Decode latent to RGBA image (outputs IMAGE + MASK)
`AlphaVAE Encode`	Encode RGBA image to latent

Installation

cd ComfyUI/custom_nodes
git clone https://github.com/katsut/ComfyUI-AlphaVAE.git
pip install diffusers>=0.33.0

Required Models

1. FLUX.1-dev (base model)

Download flux1-dev.safetensors from HuggingFace (requires license agreement).

Place in: ComfyUI/models/diffusion_models/flux1-dev.safetensors

2. AlphaVAE VAE (168 MB)

Download finetune_VAE/ directory from HuggingFace.

Place in: ComfyUI/models/vae/AlphaVAE/finetune_VAE/ (must contain config.json + diffusion_pytorch_model.safetensors)

3. AlphaVAE Diffusion LoRA (1.3 GB) — Required

Download finetune_VAE/finetune_diffusion/pytorch_lora_weights.safetensors from HuggingFace.

Place in: ComfyUI/models/loras/ (rename to e.g. alphavae_diffusion_lora.safetensors)

Important: The LoRA is required for proper transparency generation. Without it, AlphaVAE produces flat alpha values with no useful transparency information.

4. CLIP models (for FLUX text encoding)

clip_l.safetensors → ComfyUI/models/clip/ (download)
t5xxl_fp8_e4m3fn.safetensors → ComfyUI/models/clip/ (download)

Example Workflow

UNETLoader (flux1-dev.safetensors)
  → LoraLoader (alphavae_diffusion_lora.safetensors, strength: 1.0)
    → KSampler (steps: 20, cfg: 1.0, sampler: euler, scheduler: simple)

DualCLIPLoader (clip_l + t5xxl, type: flux)
  → LoraLoader (same)
    → CLIPTextEncode → FluxGuidance (guidance: 3.5)
      → KSampler

AlphaVAELoader (AlphaVAE/finetune_VAE)
  → AlphaVAEDecode (from KSampler output)
    → JoinImageWithAlpha (IMAGE + MASK → RGBA)
      → SaveImage

Requirements

ComfyUI v0.18+
diffusers >= 0.33.0

License

Code: MIT License

Model License Notice

AlphaVAE model weights are a fine-tuned derivative of FLUX.1-dev's VAE. FLUX.1-dev is released under the FLUX.1 [dev] Non-Commercial License. As a derivative work, AlphaVAE weights may inherit FLUX.1-dev's non-commercial restrictions regardless of the license stated on the AlphaVAE repository. Commercial use may require a separate license from Black Forest Labs.

This project distributes only wrapper code, not model weights. Users are responsible for obtaining model weights and complying with applicable licenses.

References

AlphaVAE Paper (arXiv:2507.09308)
AlphaVAE GitHub
AlphaVAE HuggingFace
LayerDiffuse — Prior art (SDXL, unmaintained)