[flux1-kontext-dev] Flux Kontext (image editing) model — v1.0

This Flux.1 Kontext workflow is for image editing: it can take an image and text, run the named Flux processing steps, and produce a saved image.

The listed steps cover image loading and scaling, reference-latent handling, text encoding, Flux guidance and sampling, VAE encoding and decoding, and image saving.

Use LoadImage for an image input and CLIPTextEncode for a text input.

SaveImage is the named image-output step.

Flux.1 Kontext is the base model, and Nouvo Ai is the listed publisher.

The source page and download link both point to https://civitai.com/models/1721547.

Named model files are ae.safetensors, clip_l.safetensors, flux1-kontext-dev.safetensors, and t5xxl_fp16.safetensors.

Named processing steps include BasicGuider, BasicScheduler, CLIPTextEncode, DualCLIPLoader, EmptySD3LatentImage, FluxGuidance, FluxKontextImageScale, ImageScaleBy, KSamplerSelect, and LoadImage.

Other named steps are ModelSamplingFlux, RandomNoise, ReferenceLatent, SamplerCustomAdvanced, SaveImage, UNETLoader, VAEDecode, VAEEncode, and VAELoader.

The named files are identified as follows: ae.safetensors is a VAE; flux1-kontext-dev.safetensors is a UNet; clip_l.safetensors and t5xxl_fp16.safetensors are text encoders.

Before using the workflow, verify that the named model files and processing steps are available in your environment.

Vorschlag · nicht geprüft

If it does not run, try it in Nordy.

Vorschlag · nicht geprüft