<h1 id="seemoss-genesis-ultimate">🎨 SeeMoss-Genesis Ultimate — Anime Text-to-Image Workflow for MOSS-Image (v1.0)</h1>ComfyUI workflow for SeeMoss-Genesis, my anime model on MOSS-Image-SFT (OpenMOSS, ~5B DiT). Write an idea, get a detailed anime illustration — with an optional Qwen3.5 prompt enhancer, an SFW / NSFW switch, an optional LoRA loader and two hi-res finishing branches.Every saved PNG carries Civitai-readable generation data: upload it and Civitai fills in prompt, seed, sampler and the model link by itself.How it works: 🧠 load models → 🧬 optional LoRAs → ✍️ prompt → ✨ optional enhancer → 🎨 base image → 🔬 Detail Refiner (latent upscale 1.5× + refine) and/or 🔼 Tiled Upscale (1.5×, tile by tile) → 💾 every stage saves its own PNG with Civitai metadata. Everything is grouped, color-coded, switchable and documented with notes inside the workflow, including clickable download links.<hr />
<h2 id="bounding-box-editor">📐 NEW: Bounding Box Editor version</h2>Second version on this page: Bounding Box Editor v1.0. Same models and finishing as Ultimate, but you lay out the image by drawing boxes — label them <code>character</code>, <code>object</code> or <code>text</code>, caption each one, describe the background in the summary caption, and MOSS-Image builds the scene to your layout, including headlines and typography. Defaults to <a rel="ugc" href="https://civarchive.com/models/3031179?modelVersionId=3444861">Preview-Turbo</a> (CFG 1 · 4–8 steps) with the Detail Refiner (group 07). The box nodes are ComfyUI core (recent ComfyUI needed); the custom node list below is the same. Pick the version at the top of the page — the gallery shows the same scenes as Ultimate, rebuilt with box layouts.<hr />
<h2 id="controlnet">🗺️ NEW: ControlNet version</h2>Third version on this page: ControlNet v1.0. Same models and finishing as Ultimate, plus depth control: load any reference picture and the image keeps its pose, framing and silhouette while your prompt changes character, outfit, scene and style. The depth map is made automatically (MoGe 3, ComfyUI core) and fed into the MOSS-Image color/depth control adapter by <a target="_blank" rel="ugc" href="https://huggingface.co/darask0/MOSS-Image-color-depth-control">darask0</a>. An optional Color Map (one AIO Aux Preprocessor, OneFormer-ADE20K) can also keep the background/room layout of the reference — one ON / OFF switch, default OFF. Extra downloads: <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MOSS-Image/resolve/main/loras/moss_image_color_depth_control.safetensors">control adapter</a> and <a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MoGe/resolve/main/geometry_estimation/moge_3_vitg_fp16.safetensors">MoGe 3</a>; extra custom node: comfyui_controlnet_aux (colour map only). Pick the version at the top of the page.<hr />
<h2 id="settings">⚙️ Recommended settings</h2><ul><li>🟢 Base (default): 25 steps · CFG 2.5 · Shift 5 · euler / simple</li><li>💎 Best quality: 50 steps · CFG 3 · Shift 5 · euler / simple</li><li>🚀 Turbo: CFG 1 · 4–8 steps · <a rel="ugc" href="https://civarchive.com/models/3031179?modelVersionId=3444861">Preview-Turbo</a></li><li>Skip Layer Guidance: optional, layer 22 · Triggers: seesee_elf, seesee_kitsune · pixel look: pixel art</li></ul><hr />
<h2 id="required-models">📥 Required Models</h2>All from the model page: <a rel="ugc" href="https://civarchive.com/models/3031179/seemoss-genesis">SeeMoss-Genesis on Civitai</a><ul><li>SeeMoss-Genesis (version Preview) → <code>ComfyUI/models/diffusion_models/</code> · int8_convrot (workflow default), BF16, fp8_scaled, mxfp8, nvfp4 or int4_convrot</li><li>Text encoder qwen3.5_9b_w4a8_gptq (version Text Encoder) → <code>ComfyUI/models/text_encoders/</code> · CLIP type moss_image</li><li>VAE flux2-vae_small (version VAE) → <code>ComfyUI/models/vae/</code></li><li>2x-AnimeSharpV4_RCAN by Kim2091 (optional, hi-res) → <code>ComfyUI/models/upscale_models/</code> · <a rel="ugc" href="https://github.com/Kim2091/Kim2091-Models/releases/download/2x-AnimeSharpV4/2x-AnimeSharpV4_RCAN.safetensors">Download</a></li></ul>Upstream: <a rel="ugc" href="https://huggingface.co/OpenMOSS-Team/MOSS-Image-SFT">MOSS-Image-SFT</a> · <a rel="ugc" href="https://github.com/OpenMOSS/MOSS-Image">OpenMOSS/MOSS-Image</a> · <a rel="ugc" href="https://huggingface.co/Comfy-Org/MOSS-Image">Comfy-Org repack</a> · <a rel="ugc" href="https://docs.comfy.org/tutorials/image/moss-image/moss-image">ComfyUI guide</a>⚠️ Civitai renames files on download. Open the workflow and pick your file in each loader.<hr />
<h2 id="required-custom-nodes">🧩 Required Custom Nodes</h2>Install via <a rel="ugc" href="https://github.com/Comfy-Org/ComfyUI-Manager">ComfyUI-Manager</a> (search the name) or git clone into <code>custom_nodes/</code>:<ul><li><a rel="ugc" href="https://github.com/seesee75-commits/ComfyUI-SeeSee-Civitai-Metadata">SeeSee Civitai Metadata</a> — the save nodes: Civitai-readable metadata in every PNG</li><li><a rel="ugc" href="https://github.com/willmiao/ComfyUI-Lora-Manager">ComfyUI-Lora-Manager</a> — LoRA loader + trigger words</li><li><a rel="ugc" href="https://github.com/rgthree/rgthree-comfy">rgthree-comfy</a> — hi-res group switch, image comparers</li><li><a rel="ugc" href="https://github.com/ssitu/ComfyUI_UltimateSDUpscale">ComfyUI_UltimateSDUpscale</a> — Tiled Upscale</li><li><a rel="ugc" href="https://github.com/giriss/comfy-image-saver">comfy-image-saver</a> — Seed Generator</li></ul>Everything else is ComfyUI core. Needs ComfyUI 0.42.0+ (native MOSS-Image support, subgraphs).<hr />
<h2 id="installation">🛠️ Installation</h2><ol><li>Install the 5 node packs with ComfyUI-Manager → restart ComfyUI.</li><li>Download model, text encoder, VAE (and optionally AnimeSharp) into the folders above.</li><li>Open the workflow and select your files in the loaders.</li><li>Write your idea in ✍️ Positive Prompt, set 🔞 NSFW MODE, switch groups on or off, hit Run. Images land in <code>output/</code> as <code>SeeMoss-Genesis_…</code>, <code>SeeMoss-Genesis-Refined_…</code> and <code>SeeMoss-Genesis-Tiled-Upscale_…</code>.</li></ol><hr />
<h2 id="notes">📝 Notes</h2><ul><li>✨ Prompt Enhancer: check 👁️ Final Prompt Preview before important renders. MOSS-Image likes long, natural-language English descriptions.</li><li>🔞 NSFW MODE switches the enhancer system prompt and the negative prompt together.</li><li>Adult content: consenting-adult content only, where platform and law allow.</li></ul>Credits: OpenMOSS (MOSS-Image, Apache 2.0) · Comfy-Org · Kim2091 (AnimeSharpV4) · rgthree · willmiao (LoRA Manager) · ssitu (Ultimate SD Upscale) · giriss (comfy-image-saver) · model + workflow: SeeSeeLP

SeeMoss-Genesis-Workflows

📐 SeeMoss-Genesis Bounding Box Editor v1.0A layout-first version of the SeeMoss-Genesis workflow: instead of describing where things go in words, you draw boxes on the canvas and MOSS-Image places character, objects and text inside them. Same models, same save nodes with Civitai metadata, same Detail Refiner — plus a visual 📐 Bounding Box Editor.
🧭 How to use it<ol><li>Open 📐 03 · Bounding Box Layout + Prompt and draw a box for every element in the 📐 Bounding Box Editor.</li><li>Give each box a label: <code>character</code> (a person, pose and outfit), <code>object</code> (props like a crystal, staff, table, candles) or <code>text</code> (typography).</li><li>Write a caption per box. For <code>text</code> boxes also fill in the exact words — headline, title, quote. Size and position of the box decide the size and position of the lettering.</li><li>The background has no box: describe the whole scene, setting and lighting in the summary caption.</li><li>Character boxes are passed on as <code>character_boxes</code>, so the pose and figure follow your layout. Hit Run.</li></ol>The default layout is a small example (magical girl with staff, background and a title) — just move the boxes and rewrite the captions.
🧾 MOSS layout JSONThe layout is stored in MOSS-Image's native format: a <code>summary.caption</code> for the whole scene plus a list of <code>items</code>, each with <code>label</code>, <code>caption</code>, optional <code>text</code> and a <code>bbox</code> in yxyx order with coordinates relative to the image center (<code>bbox_coord_space: abs_centralized</code>). You can paste such a JSON directly, or drag one of the gallery PNGs into ComfyUI — every image here has its workflow and box layout embedded.
⚙️ Default settings<ul><li>🚀 Turbo (default): SeeMoss-Genesis <a href="https://civitai.red/models/3031179?modelVersionId=3444861" rel="ugc">Preview-Turbo</a> int8 · CFG 1 · 4–8 steps · Shift 5 · euler / simple</li><li>🔬 Detail Refiner (group 07): latent upscale 1.5× + 8 steps · CFG 1 · denoise 0.5</li><li>Using the non-Turbo Preview model instead: 25 steps · CFG 2.5 (best quality 50 · CFG 3).</li><li>NSFW mode, Prompt Enhancer and SLG are off by default; Tiled Upscale (06) is optional and off.</li></ul>
📥 ModelsAll from <a href="https://civitai.red/models/3031179/seemoss-genesis" rel="ugc">SeeMoss-Genesis</a>: <a href="https://civitai.red/models/3031179?modelVersionId=3444861" rel="ugc">Preview-Turbo</a> (default) or Preview → <code>models/diffusion_models/</code> · Text Encoder qwen3.5_9b_w4a8_gptq (CLIP type moss_image) → <code>models/text_encoders/</code> · VAE flux2-vae_small → <code>models/vae/</code>.
🧩 Required Custom Nodes<ul><li><a href="https://github.com/seesee75-commits/ComfyUI-SeeSee-Civitai-Metadata" rel="ugc">SeeSee Civitai Metadata</a> — save nodes with Civitai-readable metadata</li><li><a href="https://github.com/willmiao/ComfyUI-Lora-Manager" rel="ugc">ComfyUI-Lora-Manager</a> — LoRA loader + trigger words</li><li><a href="https://github.com/rgthree/rgthree-comfy" rel="ugc">rgthree-comfy</a> — group switch, image comparers</li><li><a href="https://github.com/ssitu/ComfyUI_UltimateSDUpscale" rel="ugc">ComfyUI_UltimateSDUpscale</a> — Tiled Upscale</li><li><a href="https://github.com/giriss/comfy-image-saver" rel="ugc">comfy-image-saver</a> — Seed Generator</li></ul>The Bounding Box Editor and MOSS Layout Prompt nodes are ComfyUI core — no extra pack, but you need a recent ComfyUI. Update ComfyUI if the nodes show up red.
📝 Honest notes<ul><li>Boxes are a strong hint, not pixel-perfect: small objects (e.g. a quill) can come out a bit smaller, and hands don't always land exactly where you drew them.</li><li>Long text renders best in large boxes; tiny text boxes may get garbled lettering.</li><li>The gallery shows the same scenes as the Ultimate version, rebuilt with box layouts.</li></ul>

Files with SHA256: 5c5e9acaff47af71c8f96dd5e78c5196dca749c82baa9c1b1b3b6df105c5309f