Simple GGUF Hunyuan text2video workflow

This ComfyUI workflow is for Hunyuan Video text-to-video generation: it takes a text prompt, loads a GGUF model, and names an animated WEBP as the output.

The workflow covers text encoding, Hunyuan latent-video setup, advanced sampling, GGUF loading, VAE decoding, and animated WEBP saving.

Enter a text prompt through CLIPTextEncode; the named output stage is SaveAnimatedWEBP.

If ComfyUI reports a missing-node error, consider updating ComfyUI/nodes.py, ComfyUI/comfy_extras/nodes_hunyuan.py, and the GGUF node.

Vorschlag · nicht geprüft

For GGUF files, see https://huggingface.co/city96/HunyuanVideo-gguf.

Vorschlag · nicht geprüft

The workflow names these model files: clip_l.safetensors, hunyuan_video_vae_bf16.safetensors, hunyuan-video-t2v-720p-Q4_1.gguf, and llava_llama3_fp8_scaled.safetensors.

Named node types include CLIPTextEncode, DualCLIPLoader, EmptyHunyuanLatentVideo, KSamplerAdvanced, SaveAnimatedWEBP, UnetLoaderGGUF, VAEDecodeTiled, and VAELoader.

For GGUF loading, look for UnetLoaderGGUF and hunyuan-video-t2v-720p-Q4_1.gguf.

Vorschlag · nicht geprüft

For text encoding, look for CLIPTextEncode, DualCLIPLoader, clip_l.safetensors, and llava_llama3_fp8_scaled.safetensors.

Vorschlag · nicht geprüft

For VAE handling, look for VAELoader, VAEDecodeTiled, and hunyuan_video_vae_bf16.safetensors.

Vorschlag · nicht geprüft

Before use, verify that the named model files and node types are available in your ComfyUI installation.

Vorschlag · nicht geprüft

The workflow explicitly links to clip_l.safetensors; verify that file locally before use.

Vorschlag · nicht geprüft