FluxKontextDev-Lora General workflow通用工作流

Use this general FluxKontextDev-Lora workflow for the Phantasm series: provide a text prompt and reference image, then save an image output through SaveImage.

It uses Flux.1 Kontext as its base model and includes image scaling, reference-latent handling, LoRA loading, sampling, and VAE encoding and decoding.

Text is handled through CLIPTextEncode, the image and reference path uses LoadImage and ReferenceLatent, and SaveImage is listed for the image output.

The workflow is listed on Civitai under publisher Xiaolxl.

The listed model files are ae.safetensors, clip_l.safetensors, flux1-kontext-dev.safetensors, t5xxl_fp16.safetensors, and 幻象绘影HXHY亚洲风格写实-RealisticKontextLoRA.safetensors.

The listed node types include CLIPTextEncode, DualCLIPLoader, LoadImage, ReferenceLatent, EmptySD3LatentImage, FluxGuidance, FluxKontextImageScale, and ImageScaleBy.

Other listed node types are BasicGuider, BasicScheduler, KSamplerSelect, RandomNoise, SamplerCustomAdvanced, ModelSamplingFlux, UNETLoader, LoraLoaderModelOnly, VAELoader, VAEEncode, VAEDecode, and SaveImage.

For image scaling, the workflow includes FluxKontextImageScale and ImageScaleBy.

To load the listed LoRA file, the workflow includes LoraLoaderModelOnly.

Before using it, verify that ae.safetensors, clip_l.safetensors, flux1-kontext-dev.safetensors, t5xxl_fp16.safetensors, and 幻象绘影HXHY亚洲风格写实-RealisticKontextLoRA.safetensors are available in your setup.

Suggestion · not verified

Before using it, verify that your setup provides the listed nodes for text, image, reference, and image output, including CLIPTextEncode, DualCLIPLoader, LoadImage, ReferenceLatent, and SaveImage.

Suggestion · not verified