Add Hugging Face community FLUX.2-klein notes

models/flux2klein/COMMUNITY_NOTES.md

@@ -0,0 +1,51 @@
+# Hugging Face Community Notes
+
+These are useful `FLUX.2-klein` community findings from Hugging Face pages that did not all ship clean standalone workflow JSON files, but still contain setup or optimization advice worth keeping.
+
+## Unsloth GGUF discussion
+
+Source: <https://huggingface.co/unsloth/FLUX.2-klein-base-9B-GGUF/discussions/1>
+
+What looked useful:
+
+- At least one user reports a working `9B base` path by starting from the official Comfy `flux2 klein base 9b` workflow and swapping the diffusion and clip loaders to GGUF-compatible loaders.
+- The discussion explicitly calls out the text-encoder mismatch trap:
+  - `4B` uses `Qwen3 4B`
+  - `9B` uses `Qwen3 8B`
+- The thread also suggests keeping ComfyUI very current; some of the reported matrix-shape errors appear to come from version mismatch or wrong encoder pairing.
+
+Our take:
+
+- Good sign that community users are already adapting the official workflows instead of reinventing everything from scratch.
+- If you want a `9B GGUF` branch later, this discussion is one of the better breadcrumbs for why it fails and how people are fixing it.
+
+## GuangyuanSD Blitz notes
+
+Source: <https://huggingface.co/GuangyuanSD/FLUX.2-klein-9B-Blitz-ComfyUI>
+
+What looked useful:
+
+- The page is strongly oriented around low-step speed-first `9B` usage.
+- Recommended settings are very explicit:
+  - `4-5` steps
+  - `CFG = 1`
+- The author says their environment was `ComfyUI 0.11`.
+- The repo also mentions LoRA-adapter style application paths and Blackwell-oriented quantization variants.
+
+Our take:
+
+- Even though this page did not expose a clean workflow JSON in the fetched content, the settings are concrete enough to matter.
+- If you try community `9B Blitz` variants later, these are the first settings I would mirror instead of guessing.
+
+## ponpoke uncensored encoder
+
+Source: <https://huggingface.co/ponpoke/flux2-klein-9b-uncensored-text-encoder>
+
+What looked useful:
+
+- It is not a workflow, but it is directly usable in ComfyUI as a replacement text encoder placed in the `models/clip` path.
+- The author explicitly says to pair it with the official `FLUX.2-klein-9B` DiT and standard or GGUF-compatible loader nodes.
+
+Our take:
+
+- Not something to treat as a default, but worth remembering if you ever want alternate text-encoder behavior in a `9B` setup.

models/flux2klein/README.md

@@ -21,3 +21,8 @@ For outpainting specifically:
 
 - start with `image-edit-4b-distilled/` if you want the lightest and fastest first try
 - move to `image-edit-9b-base/` or `image-edit-9b-distilled/` if quality is strong enough to justify the heavier setup
+
+## Community picks on Hugging Face
+
+- `community-gguf-kv-yarvixpa/` - community `9B KV GGUF` workflow surfaced through a Hugging Face model page
+- `COMMUNITY_NOTES.md` - useful Hugging Face community optimisation notes that did not all come with clean workflow JSON files

models/flux2klein/community-gguf-kv-yarvixpa/README.md

@@ -0,0 +1,50 @@
+# Community GGUF KV 9B Workflow
+
+## Summary
+
+Community `FLUX.2-klein 9B KV GGUF` workflow linked from a Hugging Face model page. This is a good candidate if you want a more optimized community path than the pure official templates, especially around GGUF and KV-cache usage.
+
+## Model Family
+
+- `flux2klein`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Hugging Face discovery page: <https://huggingface.co/QuantStack/FLUX.2-Klein-9B-KV-GGUF>
+- Workflow file: <https://raw.githubusercontent.com/YarvixPA/ComfyUI-YarvixPA/main/example_workflows/YarvixPA%20-%20Flux2%209B%20KV%20-%20GGUF.json>
+- Workflow maintainer: `YarvixPA`
+- Hugging Face publisher that surfaced it: `QuantStack`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- The `QuantStack` Hugging Face page says this workflow is available through the author's custom node package `ComfyUI-YarvixPA`.
+- That same page says the workflow includes a `Flux KV Cache` node compatible with GGUF models and a fix that avoids limiting the editing capabilities of the model.
+- The model page also points users to `ComfyUI-GGUF` and says the model files belong in `ComfyUI/models/unet`.
+
+## Our notes
+
+- This is the strongest Hugging Face community workflow find so far because it is both `9B`-relevant and explicitly optimization-oriented.
+- The KV-cache angle is the interesting part here, not just the GGUF loader swap.
+- Good candidate when the official template works but you want a more tuned community path for local efficiency.
+
+## Required custom nodes
+
+- `ComfyUI-GGUF`
+- `ComfyUI-YarvixPA`
+- `rgthree-comfy` for `Fast Groups Bypasser`
+
+## Built-in / core nodes used
+
+- `LoadImage`
+- `SaveImage`
+- `ImageScaleToTotalPixels`
+
+## Notes
+
+- This looks more like a performance-oriented community branch than a clean beginner template.
+- Worth trying after the official `9B` templates if you want GGUF and caching-specific optimisations.