Flux-Kontext - SFW/NSFW Wan 2.2 Character Lora Training Image Generation

Overview This custom Flux-Kontext workflow is a major extension of the base model, specifically designed to handle NSFW image creation for a specific character. It builds upon the capabilities of the SFW version, offering advanced features for multi-image composition, LoRA stacking, and multi-step generation to ensure high-quality, consistent, and explicit character datasets.
Note: the lora used in this workflow is here: https://civitai.com/models/2079411?modelVersionId=2352840
Note 2: This workflow DOES support furry and anthromorphized characters - BUT you have to tell it what their tails/ears are doing in each prompt. IE. their fluffy fox tail is held out to the size...
Note 3: Do you want ALL the NSFW LoRAs, checkpoints, and Workflows? Here is my patreon - i will uploading some CRAZY stuff I have been working on.
https://www.loboforge.com/@LoboForge
Note 4: If you want my CLOTHING REMOVAL LoRA it's still available on LoboForge.
https://www.loboforge.com/@LoboForge

🖼️ Dual-Image Context System The workflow allows you to combine two different input images into a single coherent NSFW generation:

• If both inputs are active, the system blends their visual and semantic context — e.g., “the man and the woman are in an intimate embrace.”

• If only one image is active, the model evolves that image using the new prompt — e.g., “same face and clothes, the woman is in a provocative pose.”

This setup enables dynamic, instruction-based image merging without losing character identity, ensuring that explicit content is seamlessly integrated.

🧠 LoRA Power Loader A custom LoRA Power Loader lets you stack multiple LoRAs simultaneously, each with independent strength control. You can fine-tune how each LoRA influences the output, controlling aspects like:

• Facial identity

• Body type and proportions

• Clothing style (or lack thereof)

• Lighting or artistic detail to enhance explicit content

This provides granular creative direction for generating NSFW datasets or refining character details.

🔍 Latent Upscale & Restoration A built-in Latent Upscale node enhances weak or low-resolution inputs before re-generation. It restores missing structure and detail at the latent level, producing clean, high-resolution results without introducing blur or artifact noise. This is crucial for maintaining the integrity and quality of explicit content.

Use this feature to:

• Improve dataset quality for NSFW content

• Restore bad or compressed source images with explicit details

• Regenerate old renders into consistent HD-quality material with enhanced explicit elements

🧩 Multi-Prompt / Multi-Step Generation The workflow supports multiple prompts and generation passes, automatically rendering your character in:

• Different explicit poses

• Varied lighting and environments to set the mood

• Multiple angles and camera distances to capture every detail

This produces a rich, consistent dataset ideal for training LoRAs or any character model requiring 360° diversity and visual coherence in NSFW content.

🎯 Use Cases

• Creating training data for NSFW LoRA or DreamBooth

• Generating pose/lighting variation sets for the same character in explicit scenarios

• Cleaning and enhancing low-quality reference material with explicit content

• Building a consistent visual identity across NSFW scenes

⚙️ Requirements

• Flux-Kontext nodes (ComfyUI Flux / Florence integration)

• LoRA Power Loader node

• Latent Upscale node

• Any Flux-compatible base model

• Optional: additional LoRAs placed in /models/Lora/