Trending Model:#1Inklingthinkingmachines⬇16kTrending Model:#2Ternary-Bonsai-27B-ggufprism-ml⬇432kTrending Model:#3Bonsai-27B-ggufprism-ml⬇1405kTrending Model:#4Unlimited-OCRbaidu⬇2237kTrending Model:#5GLM-5.2zai-org⬇545kTrending Model:#6Qwythos-9B-Claude-Mythos-5-1M-GGUFempero-ai⬇2133kTrending Model:#7krea2-identity-editconradlocke⬇0kTrending Model:#8Qwen3.6-35B-A3B-Uncensored-HauhauCS-AggressiveHauhauCS⬇1998kTrending Model:#9Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUFDavidAU⬇63kTrending Model:#10OvisOCR2ATH-MaaS⬇17kTrending Model:#1Inklingthinkingmachines⬇16kTrending Model:#2Ternary-Bonsai-27B-ggufprism-ml⬇432kTrending Model:#3Bonsai-27B-ggufprism-ml⬇1405kTrending Model:#4Unlimited-OCRbaidu⬇2237kTrending Model:#5GLM-5.2zai-org⬇545kTrending Model:#6Qwythos-9B-Claude-Mythos-5-1M-GGUFempero-ai⬇2133kTrending Model:#7krea2-identity-editconradlocke⬇0kTrending Model:#8Qwen3.6-35B-A3B-Uncensored-HauhauCS-AggressiveHauhauCS⬇1998kTrending Model:#9Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUFDavidAU⬇63kTrending Model:#10OvisOCR2ATH-MaaS⬇17k

ComfyUI-DoRA-Dynamic-LoRA-Loader Preserves Entire Workflow Setups

A futuristic data cartridge with a glowing save icon with a central holographic preset symbol.

The ComfyUI-DoRA-Dynamic-LoRA-Loader by Xmarre now includes a State Manager that lets users save and reuse entire character and style presets containing multiple LoRA stacks, prompt templates, and settings. The loader already supported stacking DoRA and regular LoRAs with broad format compatibility and crucial fixes for Flux/Flux2 and OneTrainer exports. This new manager steps in to preserve whole workflow configurations, so you can switch between setups without manually rebuilding them every time.

Xmarre also known for ComfyUI-Spectrum-WAN-Proper and ComfyUI-DiffAid-Patches, developed the loader to handle complex DoRA loading scenarios that plain ComfyUI cannot manage, especially for Flux/Flux2 and other modern architectures. The State Manager addresses the real need for reusable character workflows by storing LoRA stacks, prompt texts, seed states, and even a reference image across sessions. By connecting a manager node to multiple DoRA loader nodes, you can save distinct stacks for face, outfit, or style and recall them instantly from a single preset.

State manager and loader compatibility

Key Features
  • Save and restore multiple DoRA loader stacks.
  • Per‑character prompt templates and seed state.
  • Auto‑strength redistribution for layer balancing.
  • Flux2 DoRA adaLN swap‑scale alignment fix.
  • Diffusers/PEFT magnitude vector normalization.
  • ZiT/Lumina2 attention QKV fusion support.
  • Slice‑aware DoRA scale for offset patches.
  • GPU‑accelerated auto‑strength analysis option.

Anyone who builds character-driven image generation workflows in ComfyUI and wants to reuse setups without reconfiguring everything by hand will find the State Manager essential. The loader’s deep compatibility features also make it a solid choice for users working with Flux/Flux2, OneTrainer, or Diffusers/PEFT DoRA models. Consumer GPU owners running local pipelines can take advantage of auto‑strength balancing and device‑aware analysis to keep quality high while controlling generation speed.

Developer notes and known behaviors

The developer cautions that auto‑strength still adds extra loader‑time computation, especially on high‑rank adapters and large backbones, but the GPU analysis path is now much faster than the older CPU‑bound method. The State Manager stores its data inside workflow widget storage and a browser backup, and future updates may bring reference images to prompt presets as well. Because the loader patches ComfyUI internals for DoRA correctness, those changes affect the entire running process, not just this node’s behavior.

"I added a State Manager to my DoRA Dynamic LoRA Loader for ComfyUI" — Source: Reddit