RedNode Studio is my custom-built creative studio for ComfyUI, designed for painters, designers, and AI artists; from first-time users to experienced workflow builders.Watch the full video tutorial:<div data-youtube-video><iframe width="640" height="480" allowfullscreen="true" autoplay="false" disablekbcontrols="false" enableiframeapi="false" endtime="0" ivloadpolicy="0" loop="false" modestbranding="false" origin playlist src="https://www.youtube.com/embed/GSwWz7awvpw?rel=1" start="0"></iframe></div> For more advanced text: <a target="_blank" rel="ugc" href="https://github.com/RedNodeAI/ComfyUI-RedNodeStudio">https://github.com/RedNodeAI/ComfyUI-RedNodeStudio</a> <hr /><h2 id="update-1.6.4">Update 1.6.4</h2><h3 id="the-changes">The Changes</h3><ul><li>A tab rail down the side, grouped with icons. Fold it, resize it, put it on either side, and pick a tab preset for the job: Basic, New image, Edit, Paint and more.</li><li>Run tab: Live, Review and Save. Watch the picture form with the pipeline and your VRAM as a chart, look back over this session's runs, and set where everything is filed.</li><li>Paint with AnyPaint. Paint over part of a picture with the normal brush, describe the whole finished picture, and the krea2_anypaint LoRA redraws just that area so it fits the scene. Everything outside the paint stays exactly as it was. About 8 seconds a paint.</li><li>Re-render: turn an illustration into a photograph, as its own tab, a Detailer pass or a Paint engine.</li><li>Tools: Upscale (SeedVR2, VOSR 2.0 or tiled, one picture or a whole folder), Re-angle, Swap and the Converter.</li><li>Hero Creator makes a clean front-on headshot out of any photo for the Identity tab.</li><li>RedNode Shelf: drop pictures on it and drag them anywhere, with five stacks.</li><li>AI tab: pictures to words on a page of its own, and Image to Text as a Detailer pass.</li></ul><h3 id="important-fixes">Important fixes</h3><ul><li>The Moodboard works properly again. With the Studio preset on custom, the Moodboard and Identity dials were being ignored and every run fell back to Balanced. If the Moodboard felt like it did nothing, this was why.</li><li>Re-render no longer fills the graphics card run after run, and a prompt imported from a picture no longer renders empty.</li></ul><hr /><h3 id="handwritten-note">Handwritten Note</h3>Hi Everyone!I've been AI painting from November 2023 and been artist for over 15 years. Daily I love the painting process of using AI. I've also hated learning comfy for years, it took me thousands of hours to understand it. The worst thing about the dam software is the look, feel and experience. I wanted to PAINT, not stare at wires and understand how to hook different VAE's into different slots. I want the BULLSHIT of Comfy to be hidden behind TOOLS that do jobs you want. That was the reason I built this.I do understand that a lot of users want to just press 1 button and go, I prefer to have full control of my images and outputs. So if you are a 1 button person you CAN do that here but its not built for it. Either way I am CHANGING it all the time and now I am making videos about it! So check out the channel and leave a like if you can. <hr /><h3 id="install-in-two-clicks-(windows)">⚡ Install in two clicks (Windows)</h3>The zip has a RedNode Setup folder. Copy its two .bat files into your ComfyUI folder (next to run_nvidia_gpu.bat, or the desktop app's base folder), close ComfyUI, run RedNode_Install_Node_Packs.bat, then RedNode_Download_Models.bat and press Enter for the recommended models. Start ComfyUI, open Templates, RedNode Studio, RedNodeStudio_V1.6.Not on Windows? Search RedNode Studio in ComfyUI Manager, load RedNode Start Here Install from the templates so Manager can install the other packs, then load RedNodeStudio_V1.6. Every model file and where it goes is on the <a target="_blank" rel="ugc" href="https://github.com/RedNodeAI/ComfyUI-RedNodeStudio">GitHub page.</a><hr /><h3 id="the-highlights">The Highlights</h3>Anime to real. Drop an illustration on Re-render and it comes back as a photograph: the same character, pose, outfit and colours, made real with the <a rel="ugc" href="https://huggingface.co/lrzjason/Anything2Real">Anything2Real LoRA.</a><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/6ff7771f-2146-4125-a20f-b42e39af3773/original=true/6ff7771f-2146-4125-a20f-b42e39af3773.jpeg" /><hr /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/82a73d1e-fb67-473b-b6e1-876941dd9d03/original=true/82a73d1e-fb67-473b-b6e1-876941dd9d03.jpeg" />Watch every render. The Run tab shows the picture forming, every stage with its time, and your VRAM live, with this session's runs down the side. Set a VRAM limit to your card and the run holds models to stay under it.<hr /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/8de1d567-a22a-41f5-8ad4-dcc3cac3d62f/original=true/8de1d567-a22a-41f5-8ad4-dcc3cac3d62f.jpeg" /> Paint what you want changed. Brush a mask straight on the picture and generate. With AnyPaint on, a new jacket or a new object fits the scene instead of looking pasted in, and nothing outside the paint changes.<hr /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/fd7459ee-d516-4bcd-9db2-3807d718e06f/original=true/fd7459ee-d516-4bcd-9db2-3807d718e06f.jpeg" />Direct the shot. Drag the camera and the subject on a top view, set height, lens and aim, place people, objects and lights, and the stage writes the camera prompt and drives the camera LoRAs.<hr /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/d260371a-cd13-44e4-8473-483d54228864/original=true/d260371a-cd13-44e4-8473-483d54228864.jpeg" />Many models, one workspace. Rigs are model setups you switch between: Krea 2, Qwen Image 2.1, Z-Image, SDXL / Illustrious / Pony or your own nodes. Draft on one, refine on another.<hr /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/bd67cd5e-427d-4b55-aa3c-7b90b94d555b/original=true/bd67cd5e-427d-4b55-aa3c-7b90b94d555b.jpeg" />Keep a face. Drop a face on Krea 2 Identity and it carries into a new scene, with a Moodboard for the style beside it.<hr /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/3738365b-bf7b-47f9-8d9b-111d2074fb84/original=true/3738365b-bf7b-47f9-8d9b-111d2074fb84.jpeg" />See the whole run before you start. The Workspace Manager lays the run out in order, opens any part in one click, and checks every node pack and model your settings need.<hr /><h2 id="requirements">📦 Requirements</h2>ComfyUI 0.27.0 or newer (0.37.0 for Qwen Image 2.1 rigs), Python 3.10 or newer, no pip dependencies.The core pack downloads nothing and runs no program, and 1.6.4 passed the Comfy Registry's automated scan.<h3 id="the-.bat-files-are-separate-helpers-you-run-yourself-and-both-are-plain-text-you-can-read-first.-(put-them-in-your-comfy-ui-install-folder)">The .bat files are separate helpers you run yourself, and both are plain text you can read first. (put them in your comfy ui install folder)</h3>I HIGHLY RECOMMEND YOU USE THE .BAT FILES. Without them you may take a while looking for the models required to run the tools.The main files, all from Comfy-Org/Krea-2 on Hugging Face:krea2_turbo_int8_convrot.safetensors (models/diffusion_models), qwen3vl_4b_fp8_scaled.safetensors (models/text_encoders),qwen_image_vae.safetensors (models/vae).Plus the Krea 2 Identity Edit LoRA by lbouaraba for faces, and the krea2_anypaint LoRA by yijunwang2 for AnyPaint. The download .bat fetches all of these; the full list, optional extras included, is on the GitHub page.Also available for Forge Neo: the Krea 2 moodboard and identity-edit toolkit.

RedNode Studio | All-in-One: Anime to Real, Paint, Camera, Multi-Model

<a target="_blank" rel="ugc" href="http://krea.ai">krea.ai</a>-style vibe transfer and identity-preserving editing for the open Krea 2 (K2) model. Ships for both platforms: a ComfyUI node pack (main download) and a matching SD WebUI Forge Neo extension set; same task, same knobs.⚡ Plug and play. These are custom nodes purpose-built for K2 — not a wrapper around a stack of helper packs. Install this one pack and the bundled example workflows run on ComfyUI core nodes alone: one node replaces your CLIPTextEncode, and that's the whole integration. No 400-node mega-workflow to untangle, no dependency shopping list. Prefer a plain UI? The same features run in Forge Neo's txt2img tab with two extensions and one patch. NEO UPDATE OCTOBER 1st Neo V3.0 releases for identity and moodboard usage. 🆕 New in v1.3: Krea 2 RedNodeOne node, the whole toolkit. Subject photo + optional scene photo + style references + your instruction in — positive AND matched negative out. Presets: balanced / max identity / style only / outfit transfer. Two-ref ordering, grounded negatives, and fit geometry are handled internally, so the classic wiring mistakes are impossible. Plug the RedNode Settings node in for full manual control with plain-language dials.👗 Outfit transfer — no prompting: drop clothing/product photos into <code>moodboard_style</code>, pick the preset — your character wears the outfit, patterns and accessories included, same face every time. For strict garment construction, also feed the garment into <code>scene_image</code> at style_strength 0.8. Multi-character scenes work too (two locked identities, one generation).Also new: RedNode Prompt Combine (growable textboxes with an order field), friendly error messages for every classic wiring mistake, and the full production pipeline Krea_Workflow_Public rebuilt on RedNode with a SAM 3 face inpainter and Detail Daemon sampler stack.Coming nextRegion-masked reference boost (boost just the face), and timestep-windowed reference influence (identity lock without composition lock). In testing now.Why not just the stock nodes?Stock ComfyUI runs Krea 2 text-to-image perfectly — this pack exists for what the stock nodes can't do with reference images. Stock image-reference encodes pass your references through as semantic description only (the model knows what's in the picture, but you get no style-vs-subject control, no strength dial, and multi-reference inputs can grid). And core's ReferenceLatent node attaches latents that K2's stock model ignores — no in-context pixel path means no true identity preservation. This pack adds both halves on top of the stock implementation: the moodboard controls and the full identity-edit recipe (in-context source latents + K2-template grounded encode + grounded negative + v1.2 fit/ref_boost), plus single-encode fusion of the two. All additive — leave the nodes unused and everything is bit-identical to stock ComfyUI.🎨 MoodboardFeed reference images — generations inherit their style/vibe. Training-free, built on K2's native Qwen3-VL vision conditioning:<ul><li>Vibe strength + extract mode (style ↔ subject: keep the palette/texture/mood, or keep the subject and let your prompt restyle it)</li><li>Style crops (scramble composition so only style transfers), indirect mode (references are hidden from the image model — structurally cannot copy poses)</li><li>Multiple references blend into one joint vibe (packed vision span — no grid/collage artifacts)</li></ul>👤 Identity EditInstruction-based editing with the community <a target="_blank" rel="ugc" href="https://civitai.com/models/2761113">krea2_identity_edit LoRA</a> (port of <a target="_blank" rel="ugc" href="https://github.com/lbouaraba/comfyui-krea2edit">ComfyUI-Krea2Edit</a>): "create a photo of this person at a night market" — same face, same outfit, relit to the new scene.<ul><li>Dual conditioning: in-context VAE tokens (RoPE frame 1) + image-grounded instruction encoding</li><li>Grounded negative (needed for CFG &gt; 1 recipes), grounding_px likeness↔obedience dial, two-ref support (scene + subject), chainable extra sources</li><li>v1.2 support: ref_boost reference-fidelity dial (try 2–6 on the subject ref) and the fit geometry — sources are fitted in pixel space to your output resolution (blur-proof, no more AR-matching requirement, matches the v1.2 LoRA's training)</li><li>Fusion: identity from the edit source + style from the moodboard in one single-encode pass</li><li>Auto face-ref prep (Forge Neo, one toggle): rebuilds your reference before editing — extracts the clean, unobstructed person (hats/glasses/hands/props removed), then builds a front-facing identity headshot. Cached per image, so it only ever runs once per reference.</li></ul>ComfyUI (main download)Unzip into <code>ComfyUI/custom_nodes/</code> (or <code>git clone</code> the repo), restart. Nodes: Krea 2 Moodboard, Moodboard Encode (packed), Identity Edit, Moodboard + Identity Fusion, Edit Source Chain, Conditioning Rebalance (per-layer 12-tap reweighting, RMS-renormalized, presets).Example workflows included: basic moodboard t2i, identity edit + moodboard fusion (v1.2 wiring), and Krea_Workflow_Public — the full production pipeline (JoyCaption auto-captioning, wildcards, fusion, AR handling, upscale pass; needs ~12 common node packs, dependency table in the GitHub README, all installable via ComfyUI-Manager).Forge Neo (additional download)<ol><li><code>git apply krea2-features-backend.patch</code> from your Forge Neo root (current neo branch)</li><li>Copy the two folders from <code>extensions/</code> into <code>extensions/</code></li></ol>Everything runs in the txt2img UI — no ComfyUI needed. Full steps in the bundled <a target="_blank" rel="ugc" href="http://INSTALL.md">INSTALL.md</a>.Shared requirements<ul><li>Text encoder: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/Krea-2">Comfy-Org/Krea-2</a> <code>qwen3vl_4b_bf16.safetensors</code> (or fp8_scaled) — the VISION variant</li><li>For editing: the identity-edit LoRA at strength 1.0 · Euler/Simple · Turbo 8 steps CFG 1 (most edits; v1.2 LoRA: 8–12 steps — 8 favors composition, 12 face detail) or Raw 20–40 steps CFG 3 (removals) · generate ≤2MP</li></ul>GitHub (full docs): ComfyUI nodes: <a target="_blank" rel="ugc" href="https://github.com/RedNodeAI/ComfyUI-Krea2Moodboard">github.com/RedNodeAI/ComfyUI-Krea2Moodboard</a> Credits<a target="_blank" rel="ugc" href="https://github.com/comfyanonymous/ComfyUI">ComfyUI</a> (Qwen3-VL ports) · <a target="_blank" rel="ugc" href="https://github.com/Haoming02">Haoming02</a> (Forge Neo) · <a target="_blank" rel="ugc" href="https://github.com/lbouaraba/comfyui-krea2edit">lbouaraba</a> (identity-edit recipe + LoRA, Apache-2.0) · nova452 &amp; huwhitememes (conditioning-rebalance mechanic, Apache-2.0) · ethanfel &amp; ostris (K2 vision-conditioning recipes) · <a target="_blank" rel="ugc" href="http://Krea.ai">Krea.ai</a> (Krea 2, Community License)Code: GPL/AGPL-3.0 per repo. Not affiliated with <a target="_blank" rel="ugc" href="http://Krea.ai">Krea.ai</a>, Haoming02, or the LoRA author. No model weights are redistributed here — download them from their linked sources.

Files with SHA256: 9a3e1dd5df4e94dd950bd74e58c4bf8ed4378f2a1b6edf61f06caa8c71cadbd6