Into the Arcane (flux redux + flux pulid) — v1.0

Use Into the Arcane as a Flux Redux and PuLID image workflow: its node names point to image and text inputs and an image saved as output.

Suggestion · non vérifié

The workflow is linked to five model files: flux1-dev-fp8.safetensors, flux1-redux-dev.safetensors, pulid_flux_v0.9.1.safetensors, sigclip_vision_patch14_384.safetensors, and t5xxl_fp8_e4m3fn.safetensors.

CLIPVisionLoader and CLIPVisionEncode appear alongside CLIPTextEncode, pointing to vision and text processing steps.

Suggestion · non vérifié

Nouvo Ai’s listing names Flux.1 D as the base model.

The listing mentions Nordy as an option if you cannot run the workflow.

The workflow also names ae.safetensors and clip_l.safetensors; check those filenames against the model files available in your setup.

Suggestion · non vérifié

Named nodes include ApplyPulidFlux, BasicGuider, BasicScheduler, CLIPTextEncode, CLIPVisionEncode, CLIPVisionLoader, DualCLIPLoader, EmptySD3LatentImage, FluxGuidance, KSamplerSelect, LoadImage, and ModelSamplingFlux.

Other named nodes are PulidFluxEvaClipLoader, PulidFluxInsightFaceLoader, PulidFluxModelLoader, RandomNoise, SamplerCustomAdvanced, SaveImage, StyleModelApply, StyleModelLoader, UNETLoader, VAEDecode, and VAELoader.

To explore guidance, sampling, noise, or style handling, look at FluxGuidance, KSamplerSelect, RandomNoise, and StyleModelApply.

Suggestion · non vérifié

No license is listed for the workflow; check the source terms before reuse.

Suggestion