--- license: apache-2.0 base_model: - Qwen/Qwen-Image-Edit-2511 - Qwen/Qwen-Image-Edit-2509 base_model_relation: finetune pipeline_tag: text-to-image tags: - text-to-image - image-to-image - image-editing - anime - qwen-image - qwen-image-edit - comfyui - safetensors - gguf - aio - checkpoint language: - en ---

# 🎨 QWEN-Anime ### Advanced anime generation with image editing — all versions on one page (Beta3-AIO is the latest)

🚀 Quick start: download Beta3-AIO (FP8), put it in ComfyUI/models/checkpoints/, load the included workflow from workflows/, then choose text-to-image or image editing (1 – 3 images). Use 4 – 8 steps, CFG 1, Euler.

## ✨ At a glance | Version | Base | Type | Files | Best for | |:--|:--|:--|:--|:--| | 💎 **Beta3-AIO** ⭐ | Qwen-Image-Edit-2511 | AIO, text-to-image + image editing | FP8 (26.99 GB) | Latest features, editing 1 – 3 images, 4+ steps | | 💎 **Beta2-AIO** | Qwen-Image-Edit-2509 | AIO | Full FP8 (quality) · Pruned FP8 (6 – 8 steps) | Plug-and-play text-to-image | | 🧠 **Beta2** | Qwen-Image-Edit-2509 | Separate VAE and text encoder | BF16 · FP16 · FP8 · GGUF (F16, Q8, Q6_K, Q4_K_S) | Flexibility, low VRAM (GGUF) | | 📦 **Beta1** (legacy) | Qwen-Image | Separate VAE and text encoder | FP16 (38.05 GB) | Old workflows, comparison | > ⚠️ All versions are combined on this model card for convenience. Details for each version follow below. > **Status (Nov 7, 2026):** The weight files (about 300 GB in total) are still being uploaded one by one. Files not yet listed under *Files and versions* will appear over time. **Jump to:** [🚀 Start here](#🚀-start-here) · [🖼️ Gallery](#🖼️-gallery) · [⬇️ Downloads](#⬇️-downloads) · [⚙️ Settings](#⚙️-settings) · [🧩 Workflows](#🧩-workflows) · [📚 More details](#📚-more-details) · [📜 License](#📜-license) ## 🚀 Start here **Newest and easiest: Beta3-AIO** 1. **Download** `Qwen-Anime-Beta3-AIO.safetensors` (FP8, about 27 GB) from [⬇️ Downloads](#⬇️-downloads) and put it in `ComfyUI/models/checkpoints/`. VAE and text encoder are inside, no extra files needed. 2. **Load the included workflow** from [🧩 Workflows](#🧩-workflows) (a standard Load Checkpoint workflow also works for text-to-image). 3. **Choose one mode and generate:** *text-to-image* (write a prompt) or *image editing* (upload 1 – 3 images, write the edit instruction). Use only one mode at a time. 4. Settings: **4 – 8 steps · CFG 1 · Euler** (fast). For more quality use up to 20 steps with CFG 3.5, Euler and Beta. **Which version should I choose?** **Choose Beta3-AIO if:** ⭐ ✅ Want image editing capabilities ✅ Need to combine multiple images ✅ Want latest features and improvements ✅ Want fastest base model (4+ steps) **Choose Beta2-AIO (Pruned) if:** ✅ Want fastest text-to-image (6-8 steps) ✅ Need quick iterations/testing ✅ Prefer simplicity (single file) ✅ 8 GB+ VRAM available **Choose Beta2-AIO (Full) if:** ✅ Want maximum quality ✅ Need more control (CFG 2.5-4.0) ✅ Creating final/detailed work ✅ Prefer traditional workflow **Choose Beta2 FP8 if:** ✅ Want flexibility (separate VAE/encoder) ✅ Using custom VAE/encoders ✅ Need maximum compatibility **Choose Beta2 GGUF if:** ✅ Limited VRAM (6-8 GB) ✅ Want smallest files (Q4 = 10 GB) ✅ CPU inference needed **Choose Beta1 if:** ✅ Compatibility with old workflows ✅ Testing/comparison purposes ## 🖼️ Gallery *All images were posted to the Beta3-AIO version on Civitai.*

## ⬇️ Downloads #💎 Beta3-AIO (latest, recommended) Latest version (Qwen Image Edit 2511 base, text-to-image + image editing, AIO). Civitai: [Qwen-Anime-Beta3-AIO](https://civitai.red/models/2122738?modelVersionId=2539173) · Workflow: [🎨 Beta3-AIO workflow](#🧩-workflows) | File | Format | ComfyUI folder | |:--|:--|:--| | [`Qwen-Anime-Beta3-AIO.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/Qwen-Anime-Beta3-AIO/Qwen-Anime-Beta3-AIO.safetensors) | FP8 AIO (VAE + text encoder integrated) | `models/checkpoints/` (Load Checkpoint) | **⚙️ Settings (from the original description):** Text-to-image / image editing: 4–8 steps · CFG 1 · Euler · Simple (Settings & Usage). Available-versions section: 4–20 steps (20 for quality) · CFG 3.5 · Euler · Beta. Workflow version: 4–8 steps · CFG 1 · Euler. ### 💎 Beta2-AIO All-in-One Beta2 (Qwen Image Edit 2509 base). Civitai: [QWEN-Anime-Beta2-AIO](https://civitai.red/models/2122738?modelVersionId=2440915) · Use the standard **Load Checkpoint** node. | File | Variant | ComfyUI folder | |:--|:--|:--| | [`qwen-anime-beta2-aio.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2-AIO/qwen-anime-beta2-aio.safetensors) | 🟢 Full model FP8 · 20+ steps · CFG 2.5–4.0 (quality mode) | `models/checkpoints/` | | [`qwen-anime-beta2-aio-8-step.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2-AIO/qwen-anime-beta2-aio-8-step.safetensors) | 🟡 Pruned model FP8 · 6–8 steps · CFG 1.0 (speed mode) | `models/checkpoints/` | **⚙️ Settings (from the original description):** Full: 20–32 steps · CFG 2.5–4.0 (sweet spot 3.6) · Euler A / Euler Normal / Beta / Simple. Pruned: 6–8 steps (optimal 8) · CFG 1.0 (max 2.0) · Euler A. *This version has no separate version description on Civitai.* ### 🧠 Beta2 (separate text encoder and VAE; safetensors and GGUF) Separate files; needs text encoder + VAE. Civitai: [QWEN-Anime-Beta2](https://civitai.red/models/2122738?modelVersionId=2427021) · Workflows: [🟦 Beta 1 and 2 (safetensors)](#🧩-workflows) · [🟨 Beta2-GGUF](#🧩-workflows) | File | Format | ComfyUI folder | |:--|:--|:--| | [`qwen-anime-Beta2-bf16.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-bf16.safetensors) | 🟪 BF16 | `models/diffusion_models/` | | [`qwen-anime-Beta2-fp16.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-fp16.safetensors) | 🟦 FP16 | `models/diffusion_models/` | | [`qwen-anime-Beta2-fp8.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-fp8.safetensors) | 🟨 FP8 | `models/diffusion_models/` | | [`qwen-anime-Beta2-fp16.gguf`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-fp16.gguf) | 🔹 F16 GGUF | `models/unet/` (ComfyUI-GGUF) | | [`qwen-anime-Beta2-Q8_0.gguf`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-Q8_0.gguf) | 🔹 Q8 GGUF | `models/unet/` | | [`qwen-anime-Beta2-Q6_K.gguf`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-Q6_K.gguf) | 🔹 Q6_K GGUF | `models/unet/` | | [`qwen-anime-Beta2-Q4_K_S.gguf`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-Q4_K_S.gguf) | 🔹 Q4_K_S GGUF | `models/unet/` | | [`qwen_2.5_vl_7b_fp8_scaled.safetensors`](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors) | Text encoder (required) | `models/text_encoders/QWEN/` | | [`qwen_image_vae.safetensors`](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors) | VAE (required) | `models/vae/QWEN/` | **⚙️ Settings (from the workflow page):** Fast mode: 4–8 steps with Lightning LoRA · CFG 1.0. Quality mode: 20–32 steps without LoRA · CFG 3.6–4.0. Samplers: Euler A, Euler Normal, Beta, Simple. ### 📦 Beta1 (legacy) Legacy release, FP16 only; needs text encoder + VAE (same as Beta2). Civitai: [QWEN-Anime-Beta1](https://civitai.red/models/2122738?modelVersionId=2401225) · Workflow: [🟦 Beta 1 and 2](#🧩-workflows) | File | Format | ComfyUI folder | |:--|:--|:--| | [`qwen-anime-v1.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta1/qwen-anime-v1.safetensors) | FP16 | `models/diffusion_models/` | **⚙️ Settings (from the workflow page):** same as Beta2 (fast 4–8 steps + Lightning LoRA · CFG 1.0; quality 20–32 steps · CFG 3.6–4.0). *This version has no version description on Civitai.* ## ⚙️ Settings #### 🎨 Beta3-AIO Settings **Text-to-Image Mode:** - Steps: 4-8 - CFG: 1 - Sampler: Euler - Scheduler: Simple - Resolution: 512×512 to 4K **Image Editing Mode:** - Steps: 4-8 - CFG: 1 - Sampler: Euler - Images: 1-3 (Image 1 required) - Tip: Higher steps for complex edits #### 🟢 Beta2-AIO Full Model (Quality Mode) - Steps: 20-32 - CFG: 2.5-4.0 (sweet spot: 3.6) - Sampler: Euler A, Euler Normal, Beta, Simple - Use for: High quality, detailed work, final renders #### 🟡 Beta2-AIO Pruned Model (Speed Mode) - Steps: 6-8 (optimal: 8) - CFG: 1.0 (max 2.0, but stay at 1.0) - Sampler: Euler A recommended - Use for: Fast iterations, testing, quick generations #### 📊 Universal Settings (All Versions) - Resolution: 512×512 to 2048×1152 - VRAM: 8 GB+ recommended - Lightning LoRAs: Compatible (4-step or 8-step) ## 🧩 Workflows Three ComfyUI workflows from [QWEN-Anime Official Workflow](https://civitai.red/models/2135240) are included in [`workflows/`](https://huggingface.co/SeeSee21/QWEN-Anime/tree/main/workflows) (original zips + extracted JSON). Each block shows the real workflow overview, what it does, settings, required models with download links and ComfyUI folders, and custom nodes. Result images are folded below each workflow; they come from the respective workflow version on Civitai.

🧩 QWEN-Anime-Beta3-AIO workflow

**Download:** [`workflows/Qwen-Anime-Beta3-AIO-workflow.zip`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/workflows/Qwen-Anime-Beta3-AIO-workflow.zip) (original) · [`workflows/json/Qwen-Anime-Beta3-AIO-workflow.json`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/workflows/json/Qwen-Anime-Beta3-AIO-workflow.json) · **Civitai:** [QWEN-Anime-Beta3-AIO](https://civitai.red/models/2135240?modelVersionId=2540517) ![🎨 QWEN-Anime-Beta3-AIO workflow · workflow overview](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/workflows/beta3_overview.jpg) **What it does:** Image editing (1–3 images) + text-to-image in one workflow; switch the mode with the 📐 Image Size / 🖼️ Image Editing toggles (only one mode at a time). **⚙️ Settings:** 4–8 steps · CFG 1 · Euler · tested on RTX 4060 8 GB. **📥 Required models** | File | Required? | Download | ComfyUI folder | |:--|:--|:--|:--| | Checkpoint (`CheckpointLoaderSimple`) | required | [`Qwen-Anime-Beta3-AIO.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/Qwen-Anime-Beta3-AIO/Qwen-Anime-Beta3-AIO.safetensors) (AIO: VAE + text encoder included) | `models/checkpoints/` | **🧩 Required custom nodes** | Custom node pack (external) | Nodes used in this workflow | Required? | |:--|:--|:--| | [rgthree-comfy](https://github.com/rgthree/rgthree-comfy) | `Fast Groups Bypasser`, `Any Switch`, `Lora Loader Stack`, `Image Comparer` | required | | [ComfyUI-Easy-Use](https://github.com/yolain/ComfyUI-Easy-Use) | `easy positive`, `easy negative` | required | Install with [ComfyUI-Manager](https://github.com/Comfy-Org/ComfyUI-Manager) (search the name) or `git clone` into `custom_nodes/`, then restart ComfyUI. Everything else is ComfyUI core. The Civitai text says "3 Custom Nodes" but lists two packs; the workflow JSON uses exactly these two. Image editing uses the core `TextEncodeQwenImageEditPlus` / `ReferenceLatent` nodes.

🖼️ 🎨 QWEN-Anime-Beta3-AIO workflow · result images (2)

📄 🎨 QWEN-Anime-Beta3-AIO workflow · full Civitai version description

🎨 QWEN-Anime-Beta3-AIO Workflow **Image Editing + Text-to-Image in one workflow!** 🚀 NEW in Beta3 ✨ **Dual Mode:** * 📝 Text-to-Image - Pure text generation * 🎨 Image Editing - Edit/combine 1-3 images ⚠️ **One mode at a time! Both = error.** ⚙️ Requirements **3 Custom Nodes:** 1. rgthree-comfy 2. ComfyUI-Easy-Use **Model:** Qwen-Anime-Beta3-AIO.safetensors 🎯 Settings * Steps: 4-8 * CFG: 1 * Sampler: Euler * Tested: RTX 4060 8 GB ✅ 💡 How to Use **Text-to-Image:** 1. Enable 📐 Image Size toggle 2. Disable 🖼️ Image Editing toggle 3. Write prompt → Generate **Image Editing:** 1. Enable 🖼️ Image Editing toggle 2. Disable 📐 Image Size toggle 3. Upload 1-3 images → Write prompt → Generate 🎨 Image Editing Can: * Combine characters from multiple images * Merge scenes and backgrounds * Transform styles and compositions * Change details, clothing, colors ❓ FAQ Q: Both modes at once? A: **NO!** Only ONE mode. Both = error. Q: How many images? A: 1-3 (Image 1 required) Q: Works with Beta2? A: No, Beta3 only. **Download → Install nodes → Load → Choose mode → Generate!** Created with ❤️ for QWEN-Anime community

🧩 QWEN-Anime-Beta2-GGUF workflow

**Download:** [`workflows/QWEN-Anime-Beta2-GGUF-Workflow-v2.zip`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/workflows/QWEN-Anime-Beta2-GGUF-Workflow-v2.zip) (original) · [`workflows/json/QWEN-Anime-Beta2-GGUF-Workflow-v2.json`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/workflows/json/QWEN-Anime-Beta2-GGUF-Workflow-v2.json) · **Civitai:** [QWEN-Anime-Beta2-GGUF](https://civitai.red/models/2135240?modelVersionId=2435249) ![🟨 QWEN-Anime-Beta2-GGUF workflow · workflow overview](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/workflows/beta2-gguf_overview.jpg) **What it does:** Text-to-image for the GGUF files of Beta2 (`UnetLoaderGGUF`), with optional Lightning LoRA and Hires Fix upscaler with comparison. **⚙️ Settings:** Fast: 4–8 steps with Lightning LoRA · CFG 1.0. Quality: 20–32 steps · CFG 3.6–4.0 (workflow page). **📥 Required models** | File | Required? | Download | ComfyUI folder | |:--|:--|:--|:--| | GGUF model (`UnetLoaderGGUF`; default `qwen-anime-Beta2-Q4_K_S.gguf`) | required | [`qwen-anime-Beta2-Q4_K_S.gguf`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta2/qwen-anime-Beta2-Q4_K_S.gguf) · other GGUF files in the Beta2 table above | `models/unet/` | | Text encoder (`CLIPLoader`) | required | [`qwen_2.5_vl_7b_fp8_scaled.safetensors`](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors) | `models/text_encoders/QWEN/` | | VAE (`VAELoader`) | required | [`qwen_image_vae.safetensors`](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors) | `models/vae/QWEN/` | | Lightning LoRA (`LoraLoader`; the workflow loads `Qwen\Qwen-Image-Edit-2509-Lightning-8steps-V1.0-bf16.safetensors`) | optional | [`Qwen-Image-Edit-2509-Lightning-8steps-V1.0-bf16.safetensors`](https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Edit-2509/Qwen-Image-Edit-2509-Lightning-8steps-V1.0-bf16.safetensors) (lightx2v) · links in the workflow notes: [8-Step](https://civitai.com/models/2046883) · [4-Step](https://civitai.com/models/2048940) | `models/loras/Qwen/` | **🧩 Required custom nodes** | Custom node pack (external) | Nodes used in this workflow | Required? | |:--|:--|:--| | [rgthree-comfy](https://github.com/rgthree/rgthree-comfy) | `Fast Groups Bypasser`, `Image Comparer` | required | | [ComfyUI-GGUF](https://github.com/city96/ComfyUI-GGUF) | `UnetLoaderGGUF` | required | Install with [ComfyUI-Manager](https://github.com/Comfy-Org/ComfyUI-Manager) (search the name) or `git clone` into `custom_nodes/`, then restart ComfyUI. Everything else is ComfyUI core. The workflow page also lists comfyui-copilot for both workflows, but this GGUF JSON does not use a copilot node (no `SDXLResolutionPresets`). *This workflow version has no separate version description on Civitai.*

🧩 QWEN-Anime-Beta 1 and 2 workflow

**Download:** [`workflows/QWEN-Anime-Beta1-Workflow.zip`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/workflows/QWEN-Anime-Beta1-Workflow.zip) (original) · [`workflows/json/QWEN-Anime-Beta1-Workflow.json`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/workflows/json/QWEN-Anime-Beta1-Workflow.json) · **Civitai:** [QWEN-Anime-Beta 1 and 2](https://civitai.red/models/2135240?modelVersionId=2415389) ![🟦 QWEN-Anime-Beta 1 and 2 workflow · workflow overview](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/workflows/beta1-2_overview.jpg) **What it does:** Text-to-image for the Beta1 / Beta2 safetensors (`UNETLoader`), with resolution presets, optional Lightning LoRA and Hires Fix upscaler with side-by-side comparison. **⚙️ Settings:** Fast: 4–8 steps with Lightning LoRA · CFG 1.0. Quality: 20–32 steps · CFG 3.6–4.0. Hires Fix 1.5× · denoise 0.55. **📥 Required models** | File | Required? | Download | ComfyUI folder | |:--|:--|:--|:--| | Diffusion model (`UNETLoader`; default `qwen-anime-v1.safetensors`) | required | [`qwen-anime-v1.safetensors`](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/QWEN-Anime-Beta1/qwen-anime-v1.safetensors) · or a Beta2 safetensors file from the table above | `models/diffusion_models/` | | Text encoder (`CLIPLoader`) | required | [`qwen_2.5_vl_7b_fp8_scaled.safetensors`](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors) | `models/text_encoders/QWEN/` | | VAE (`VAELoader`) | required | [`qwen_image_vae.safetensors`](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors) | `models/vae/QWEN/` | | Lightning LoRA (`LoraLoader`; the workflow loads `Qwen\Qwen-Image-Edit-2509-Lightning-8steps-V1.0-bf16.safetensors`) | optional | [`Qwen-Image-Edit-2509-Lightning-8steps-V1.0-bf16.safetensors`](https://huggingface.co/lightx2v/Qwen-Image-Lightning/resolve/main/Qwen-Image-Edit-2509/Qwen-Image-Edit-2509-Lightning-8steps-V1.0-bf16.safetensors) (lightx2v) · links in the workflow notes: [8-Step](https://civitai.com/models/2046883) · [4-Step](https://civitai.com/models/2048940) | `models/loras/Qwen/` | **🧩 Required custom nodes** | Custom node pack (external) | Nodes used in this workflow | Required? | |:--|:--|:--| | [rgthree-comfy](https://github.com/rgthree/rgthree-comfy) | `Fast Groups Bypasser`, `Image Comparer` | required | | [comfyui-copilot](https://github.com/hylarucoder/comfyui-copilot) | `SDXLResolutionPresets` | required | Install with [ComfyUI-Manager](https://github.com/Comfy-Org/ComfyUI-Manager) (search the name) or `git clone` into `custom_nodes/`, then restart ComfyUI. Everything else is ComfyUI core.

🖼️ 🟦 QWEN-Anime-Beta 1 and 2 workflow · result images (9)

📄 🟦 QWEN-Anime-Beta 1 and 2 workflow · full Civitai version description

🎨 Official ComfyUI Workflow for QWEN-Anime-Beta1 and Beta2 This is the optimized workflow designed specifically for QWEN-Anime-Beta1 checkpoint. It includes everything you need to get started quickly and generate high-quality anime images. ✨ Features: - Step-by-step workflow with clear color-coded groups - Integrated model download links and folder structure guide - Support for Lightning LoRAs (4-step and 8-step) for fast generation - Optional Hires Fix upscaler with side-by-side comparison - Preset resolutions from 512×512 to 2K (2048×1152) - Works on 8 GB VRAM 📦 What's Included: - Complete model loading setup - Prompt templates with tips - Resolution presets (SDXLResolutionPresets) - Base generation + optional upscaling - Image comparison tool (rgthree) - All necessary nodes with helpful titles and descriptions ⚙️ Requirements: - ComfyUI - Custom Nodes: rgthree-comfy, comfyui-copilot (installation links included in workflow) - QWEN-Anime-Beta1 checkpoint + VAE + Text Encoder (download links in workflow) - Optional: Lightning LoRAs for speed boost 🚀 Quick Start: 1. Load the workflow in ComfyUI 2. Follow the Model Links note to download required files 3. Place files in correct folders (structure shown in workflow) 4. Write your prompts 5. Select resolution 6. Generate! Perfect for both beginners and advanced users. All model links, installation instructions, and tips are built into the workflow itself.

📄 Full description of the Civitai workflow page (QWEN-Anime Official Workflow)

#### 🎨 Official ComfyUI Workflows for QWEN-Anime-Beta1, Beta2 & Beta3 **Two optimized workflows** for different model formats - choose what fits your setup best! **NEW: Image Editing + Text-to-Image in one workflow!** --- #### 🚀 What's New in Beta3 ✨ **Dual Mode Workflow:** * 📝 Text-to-Image mode * 🎨 Image Editing mode (NEW!) * Edit 1-3 images simultaneously * Combine characters, merge scenes, transform styles --- #### ✨ Features 🎯 **Core Capabilities:** * ✅ **Image Editing** - Edit/combine 1-3 images (NEW!) * ✅ **Text-to-Image** - Pure text generation * ✅ Toggle-based mode switching * ✅ Color-coded workflow groups * ✅ Built-in instructions and tips * ✅ Resolution presets (512×512 to 4K) * ✅ Optional upscaler with comparison * ✅ Metadata auto-save --- #### 📐 Workflow Steps 1. **📦 Load Models** - Checkpoint + optional LoRA 2. **🖼️ Upload Images** - For editing mode (1-3 images) 3. **📐 Set Size** - For text-to-image mode 4. **✍️ Write Prompts** - Positive + negative 5. **🎨 Generate** - Create your image 6. **🔼 Upscale** - Optional quality boost --- #### 🎯 Two Modes ##### 📝 Text-to-Image Mode: * Enable: 📐 Image Size toggle * Disable: 🖼️ Image Editing toggle * Pure text-based generation ##### 🎨 Image Editing Mode (NEW!): * Enable: 🖼️ Image Editing toggle * Disable: 📐 Image Size toggle * Upload 1-3 images * Combine, transform, edit ⚠️ **CRITICAL:** Only enable ONE mode at a time! --- #### ⚙️ Requirements **Custom Nodes (3):** * ✅ **rgthree-comfy** - * ✅ **ComfyUI-Easy-Use** - **Model Files:** * 🎨 Qwen-Anime-Beta3-AIO.safetensors --- #### 📦 Available Workflows for Beta1 and Beta2 ##### 🟦 Standard Workflow (Safetensors) **For:** BF16, FP16, FP8 versions **Best for:** Most users with 8 GB+ VRAM **Model:** [QWEN-Anime-Beta2 (All Versions)](https://civitai.com/models/2122738?modelVersionId=2427021) ##### 🟨 GGUF Workflow (Quantized) **For:** Q4, Q6, Q8, F16 GGUF versions **Best for:** Limited VRAM or CPU inference **Requires:** ComfyUI-GGUF custom node **Model:** [QWEN-Anime-Beta2 GGUF](https://civitai.com/models/2122738?modelVersionId=2427021) --- #### ✨ Features (Both Workflows) ##### 🎯 Core Features ✅ **Color-coded workflow groups** - Easy to follow step-by-step ✅ **Built-in model links** - Download links right in the workflow ✅ **Folder structure guide** - Never wonder where files go ✅ **Resolution presets** - 512×512 to 2K (2048×1152) ✅ **Lightning LoRA support** - 4-step & 8-step for speed ✅ **Hires Fix upscaler** - Optional 1.5x quality boost ✅ **Image comparison** - Side-by-side before/after ##### 📐 Organized Workflow Steps 1. **📦 Load Models** (Blue) - Checkpoint, VAE, Text Encoder, LoRA 2. **✍️ Write Prompts** (Green) - Positive & negative with tips 3. **📐 Choose Size** (Cyan) - Quick resolution presets 4. **🎨 Generate** (Purple) - KSampler creates your image 5. **🔼 Upscale** (Orange) - Optional quality enhancement --- #### 🔧 What's Included ##### In Every Workflow: * 📝 **Prompt templates** with quality tags and tips * 🎛️ **Resolution presets** via SDXLResolutionPresets node * 🔀 **Base generation** + optional upscaling path * 🔍 **Image comparer** (rgthree) for quality comparison * 💡 **Helpful node titles** - Every node is labeled clearly * 📚 **Complete instructions** - All info built into workflow ##### Workflow-Specific: **Standard Workflow:** * Uses `UNETLoader` node * Files in `diffusion_models/` folder * Works with BF16/FP16/FP8 formats **GGUF Workflow:** * Uses `UnetLoaderGGUF` node * Files in `unet/` folder * Requires ComfyUI-GGUF custom node * Smaller file sizes (10-20 GB) --- #### ⚙️ Requirements ##### Both Workflows Need: * ✅ ComfyUI (latest version) * ✅ **rgthree-comfy** - Image comparison & group toggle * ✅ **comfyui-copilot** - Resolution presets ##### GGUF Workflow ALSO Needs: * ⚠️ **ComfyUI-GGUF** - ESSENTIAL for GGUF models! ##### Model Files (Download from model page): * 🎨 **Checkpoint** - Beta1 or Beta2 (your chosen format) * 🧠 **Text Encoder** - qwen\_2.5\_vl\_7b\_fp8\_scaled.safetensors * 🖼️ **VAE** - qwen\_image\_vae.safetensors * ⚡ **Lightning LoRA** (Optional) - 4-step or 8-step *All download links are included in the workflows!* --- #### 🚀 Quick Start Guide ##### 1️⃣ Choose Your Workflow * **Standard** → If using BF16/FP16/FP8 safetensors * **GGUF** → If using Q4/Q6/Q8/F16 GGUF models ##### 2️⃣ Install Requirements ``` Standard: ComfyUI + rgthree + copilot GGUF: ComfyUI + rgthree + copilot + ComfyUI-GGUF ``` ##### 3️⃣ Download Model Files Visit the model page and download: * Your chosen checkpoint format * Text Encoder + VAE (both workflows use same ones) * Optional: Lightning LoRA for speed ##### 4️⃣ Place Files Correctly **Standard Workflow:** ``` 📂 ComfyUI/models/ ├── diffusion_models/ → checkpoint here ├── loras/Qwen/ → Lightning LoRA here ├── vae/QWEN/ → VAE here └── text_encoders/QWEN/ → Text Encoder here ``` **GGUF Workflow:** ``` 📂 ComfyUI/models/ ├── unet/ → GGUF checkpoint here ⚠️ ├── loras/Qwen/ → Lightning LoRA here ├── vae/QWEN/ → VAE here └── text_encoders/QWEN/ → Text Encoder here ``` ##### 5️⃣ Load & Generate! 1. Drag workflow JSON into ComfyUI 2. Check the **Model Links** note in workflow 3. Write your prompts 4. Select resolution 5. Click Queue Prompt 6. Create amazing anime art! 🎨 --- #### 💡 Which Workflow Should I Use? ##### Use **Standard Workflow** if: ✅ You have 8 GB+ VRAM ✅ Using BF16, FP16, or FP8 safetensors ✅ Want maximum quality ✅ Don't mind larger file sizes (19-38 GB) ##### Use **GGUF Workflow** if: ✅ You have limited VRAM (under 8 GB) ✅ Want smaller file sizes (10-20 GB) ✅ Need CPU inference support ✅ Storage/bandwidth limited ✅ Don't mind installing one extra custom node **Both workflows produce excellent results!** GGUF is just more memory-efficient. --- #### 🎯 Performance Comparison Feature Standard GGUF Q8 GGUF Q4 File Size 19-38 GB 20 GB 10 GB VRAM Usage 8 GB+ 7-8 GB 6-7 GB Quality Maximum Near-Max Good Speed Fast Fast Faster Setup Simple +1 Node +1 Node --- #### 📖 Workflow Features in Detail ##### Prompt System * **Positive Prompt** with quality tag examples * **Negative Prompt** with common artifacts to avoid * Built-in prompting tips and structure guide ##### Resolution Management * Quick presets via dropdown menu * Tested resolutions: 512×512 up to 2048×1152 * Custom resolution support * All major aspect ratios covered ##### Generation Settings * **Fast Mode:** 4-8 steps with Lightning LoRA (CFG 1.0) * **Quality Mode:** 20-32 steps without LoRA (CFG 3.6-4.0) * All confirmed samplers: Euler A, Euler Normal, Beta, Simple ##### Upscaler Module (Optional) * Toggle ON/OFF with one click * 1.5x latent upscale * Hires Fix refinement pass * Side-by-side comparison tool * Denoise: 0.55 for balanced refinement --- #### 🎨 Example Settings ##### Fast Generation (with 8-step LoRA): * **Steps:** 8 * **CFG:** 1.0 * **Resolution:** 512×512 or 1024×1024 * **Time:** ~20-70 seconds ##### High Quality (no LoRA): * **Steps:** 24-32 * **CFG:** 3.6-4.0 * **Resolution:** 1024×1536 * **Upscaler:** ON * **Time:** ~200-400 seconds --- #### 🔍 What Makes These Workflows Special? ##### vs Basic Setup: ✅ **Everything included** - No guessing about settings ✅ **Visual organization** - Color-coded groups ✅ **Built-in guidance** - Tips and instructions in workflow ✅ **Optimized nodes** - Pre-configured for best results ##### vs Other Workflows: ✅ **GGUF support** - Separate optimized GGUF workflow ✅ **Beginner-friendly** - Clear labels and structure ✅ **Advanced features** - Upscaler, comparison, presets ✅ **Complete documentation** - All links and info included --- #### 📚 Additional Resources ##### Model Page: **QWEN-Anime-Beta2:** ##### Custom Nodes: * **rgthree-comfy:** * **comfyui-copilot:** * **ComfyUI-GGUF:** ##### Files to Download: * **Text Encoder:** [qwen\_2.5\_vl\_7b\_fp8\_scaled.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/text_encoders/qwen_2.5_vl_7b_fp8_scaled.safetensors) * **VAE:** [qwen\_image\_vae.safetensors](https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors) * **Lightning LoRAs:** [4-Step](https://civitai.com/models/2048940) | [8-Step](https://civitai.com/models/2046883) --- #### ❓ FAQ **Q: Do I need both workflows?** A: No! Choose based on your model format. Safetensors → Standard. GGUF → GGUF Workflow. **Q: Can I use GGUF with the standard workflow?** A: No, GGUF requires the special GGUF workflow with UnetLoaderGGUF node. **Q: Which is better quality?** A: Standard (FP8/BF16) has slightly higher quality. GGUF Q8 is very close. GGUF Q4 is still good. **Q: Do I need Lightning LoRAs?** A: No, but highly recommended for speed! Without them, use 20-32 steps instead. **Q: Will this work with Beta1?** A: Yes! Both workflows work with Beta1 and Beta2. **Q: Can I add my own LoRAs?** A: Absolutely! These are full workflows - customize as you like. **Q: What if nodes are missing?** A: Install via ComfyUI Manager or follow the GitHub links above. **Q: Can I use both modes?** A: No! Only ONE mode at a time. Both enabled = error. **Q: How many images can I edit?** A: 1-3 images (Image 1 required, 2-3 optional). **Q: Which mode for pure generation?** A: Text-to-Image mode (disable image editing). **Q: Does it work with Beta2 models?** A: No, Beta3 workflow requires Beta3-AIO model. --- #### 🎯 Perfect For * 🆕 **Beginners** - Clear step-by-step workflow * 🎨 **Artists** - Focus on creativity, not technical setup * ⚡ **Speed Users** - Lightning LoRA support * 💾 **Limited Storage** - GGUF workflow available * 🔧 **Advanced Users** - Full control and customization * 📱 **Low VRAM** - GGUF Q4 works on 6-7 GB --- #### 🚀 Download & Get Started 1. **Download the workflow** you need (Standard or GGUF) 2. **Install required custom nodes** via ComfyUI Manager 3. **Download model files** from the model page 4. **Follow the built-in instructions** in the workflow 5. **Start creating** amazing anime art! Both workflows include everything you need to get started - no external guides required! --- **Created with ❤️ for the QWEN-Anime community** *Compatible with Beta1 and Beta2 | Tested on 8 GB VRAM | Works on ComfyUI latest version* --- #### 📦 Download Files **Standard Workflow:** `QWEN-Anime-Beta2-Workflow.json` **GGUF Workflow:** `QWEN-Anime-Beta2-GGUF-Workflow.json` Both included in this upload!

## 📚 More details *Everything below is collapsed. Click a title to open it.*

📄 Beta3-AIO · full Civitai description and gallery

🎨 VERSION 3 - MAJOR UPDATE: IMAGE EDITING REVOLUTION WHAT'S NEW: 📸 IMAGE EDITING FUNCTIONALITY: - Edit and transform 1-3 images simultaneously - Combine characters, scenes, and styles - Image-to-Image + Text-to-Image dual workflow - Advanced composition and modification capabilities ⚡ PERFORMANCE IMPROVEMENTS: - Base model upgraded: Qwen Image Edit 2511 (from 2509) - Minimum steps reduced: 4 steps (down from 8) - Slightly faster generation speed - Optimized for efficiency 🔓 ENHANCED CREATIVE FREEDOM: - Custom Qwen 2.5 VL 7B FP8 uncut model - Maximum creative possibilities 📦 FORMAT & AVAILABILITY: - AIO version: FP8 only - Other formats available on request 💡 TIPS FOR BEST RESULTS: - Results depend on seed, prompt, and input images - Experiment with different combinations EVOLUTION: From pure Text-to-Image model to advanced Image-to-Image editing with full Text-to-Image support! *One block of the original version description about adult content is not reproduced here.*

🖼️ Beta2-AIO · gallery

#### 🖼️ QWEN-Anime-Beta2-AIO gallery All images below were posted to the **QWEN-Anime-Beta2-AIO** version on Civitai.

📄 Beta2 · full Civitai description and gallery

📦 **Here's what each file contains:** 🧠 SafeTensor Versions: 🟪 **Pruned Model bf16 (38.05 GB)** = BF16 version 🟦 **Pruned Model fp16 (38.05 GB)** = FP16 version 🟨 **Pruned Model fp8 (19.03 GB)** = **FP8 version** 💾 GGUF Versions: *(Requires [ComfyUI-GGUF](https://github.com/city96/ComfyUI-GGUF).)* 🔹 **Full Model fp16 (38.07 GB)** = F16 GGUF 🔹 **Full Model fp8 (20.23 GB)** = Q8 GGUF 🔹 **Full Model bf16 (15.63 GB)** = Q6\_K GGUF 🔹 **Full Model nf4 (10.72 GB)** = Q4\_K\_S GGUF #### 🖼️ QWEN-Anime-Beta2 gallery All images below were posted to the **QWEN-Anime-Beta2** version on Civitai.

🖼️ Beta1 · gallery

#### 🖼️ QWEN-Anime-Beta1 gallery All images below were posted to the **QWEN-Anime-Beta1** version on Civitai.

📢 Version updates

### 🎨 VERSION 3 - LATEST (Beta3-AIO) #### 🚀 MAJOR UPDATE - Image Editing Revolution! **NEW FEATURES:** - ✨ **Image Editing functionality** - Edit 1-3 images simultaneously - 🔄 **Dual workflow** - Text-to-Image AND Image-to-Image - 📦 **Upgraded base model** - Qwen Image Edit 2511 (from 2509) - ⚡ **Faster generation** - 4 steps minimum (down from 8) - 🔓 **Custom uncut model** - Qwen 2.5 VL 7B FP8 text encoder - 📦 **FP8 only** - Other formats available on request **IMPROVEMENTS:** - Combine multiple characters from different images - Transform and merge scenes - Style transfer between images - Enhanced detail preservation - More consistent results **TIPS FOR BEST RESULTS:** - Results depend on seed, prompt, and input images - Experiment with different combinations **WORKFLOW EVOLUTION:** - V1/V2: Text-to-Image only - V3: Text-to-Image + Image-to-Image ⭐ ### 🔄 VERSION 2 (Beta2-AIO) **FEATURES:** - All-in-One format (no separate VAE/Text Encoder needed) - Two variants: Full (20+ steps) and Pruned (6-8 steps) - FP8 precision (26.99 GB) - Integrated VAE + Text Encoder - Single file, plug-and-play **IMPROVEMENTS:** - Easier setup vs Beta1 - Same quality, simpler workflow - Lightning LoRA compatible ### 📦 VERSION 1 (Beta1 - Legacy) **FEATURES:** - Original release - FP16 only (38.05 GB) - Requires separate VAE + Text Encoder - Base training | | BETA3-AIO ⭐ | BETA2-AIO | BETA2 | BETA1 | | :--- | :---: | :---: | :---: | :---: | | Image Editing | ✅ 1-3 imgs | ❌ No | ❌ No | ❌ No | | Text-to-Image | ✅ Yes | ✅ Yes | ✅ Yes | ✅ Yes | | Base Model | 2511 | 2509 | 2509 | Base | | Min Steps | 4 | 6-8 | 6-8 | 20+ | | Setup | Single | Single | 3 file | 3 file | | VAE/Encoder | Integrated | Integrated | Separate | Separate | | File Size | 27 GB | 27 GB | 19-38 | 38 GB | | Format | FP8 | FP8 | Multi | FP16 | | Speed (8 steps) | ⚡⚡⚡ | ⚡⚡⚡ | ⚡⚡ | ⚡ |

🎯 Available versions and formats

### 💎 Beta3-AIO (Recommended) ⭐ **Format:** - 🟡 FP8 (26.99 GB) - 4+ steps, CFG 3.5 - Other formats: Available on request **What's included:** - ✅ Image Editing (1-3 images) - ✅ Text-to-Image - ✅ Integrated VAE + Text Encoder - ✅ Uncut text encoder for creative freedom **Settings:** - Steps: 4-20 (recommend 20 for quality) - CFG: 3.5 - Sampler: Euler - Scheduler: Beta ### 💎 Beta2-AIO **Variants:** - 🟢 Full Model FP8 (26.99 GB) - 20+ steps, CFG 2.5-4.0 (quality mode) - 🟡 Pruned Model FP8 (26.99 GB) - 6-8 steps, CFG 1.0 (speed mode) **What's included:** - ✅ Integrated VAE + Text Encoder - ✅ Single file, plug-and-play - ✅ Use regular "Load Checkpoint" node ### 🧠 Beta2 (Safetensors & GGUF) **Requires separate VAE + Text Encoder** **SafeTensor Versions:** - 🟪 BF16 (38.05 GB) - 🟦 FP16 (38.05 GB) - 🟨 FP8 (19.03 GB) **GGUF Versions:** ⚠️ Requires ComfyUI-GGUF - 🔹 F16 (38.07 GB) - 🔹 Q8 (20.23 GB) - 🔹 Q6_K (15.63 GB) - 🔹 Q4_K_S (10.72 GB) **Version description (Civitai) – what each file contains:** - 🧠 SafeTensor: 🟪 Pruned Model bf16 (38.05 GB) = BF16 version · 🟦 Pruned Model fp16 (38.05 GB) = FP16 version · 🟨 Pruned Model fp8 (19.03 GB) = FP8 version - 💾 GGUF (requires [ComfyUI-GGUF](https://github.com/city96/ComfyUI-GGUF)): 🔹 Full Model fp16 (38.07 GB) = F16 GGUF · 🔹 Full Model fp8 (20.23 GB) = Q8 GGUF · 🔹 Full Model bf16 (15.63 GB) = Q6_K GGUF · 🔹 Full Model nf4 (10.72 GB) = Q4_K_S GGUF ### 📦 Beta1 (Legacy) - FP16 only (38.05 GB) - Requires separate VAE + Text Encoder ### Files in this repository | Folder | Civitai version | Files | | :--- | :--- | :--- | | `Qwen-Anime-Beta3-AIO/` | Qwen-Anime-Beta3-AIO | `Qwen-Anime-Beta3-AIO.safetensors` (FP8 AIO) | | `QWEN-Anime-Beta2-AIO/` | QWEN-Anime-Beta2-AIO | `qwen-anime-beta2-aio.safetensors` (Full), `qwen-anime-beta2-aio-8-step.safetensors` (Pruned) | | `QWEN-Anime-Beta2/` | QWEN-Anime-Beta2 | `qwen-anime-Beta2-{bf16, fp16, fp8}.safetensors`, `qwen-anime-Beta2-{fp16, Q8_0, Q6_K, Q4_K_S}.gguf` | | `QWEN-Anime-Beta1/` | QWEN-Anime-Beta1 | `qwen-anime-v1.safetensors` (FP16) |

🧪 Test results

### 🎨 Beta3-AIO Image Editing Test Tested on Nvidia RTX 4060 with Euler sampler ![Beta3-AIO image editing steps](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta3_edit_steps.jpg) **Test: Multi-Image Composition** **Prompt:** ```text Place the two figures in a fantasy medieval tavern, laughing and clinking two beer glasses. ``` ![Beta3-AIO multi-image composition](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta3_multi_image_composition.jpg) **Images Used:** - Image 1: Character A - Image 2: Character B **Result:** - Successfully combined both characters - Tavern setting accurately generated - Natural interaction and poses - Consistent anime style maintained ### 📊 Beta2-AIO Test Results Tested on Nvidia RTX 4060 with Euler A sampler #### 🟢 Full Model fp8 (20+ Steps Version) **Test 1: Elegant Shrine Maiden** **Resolution:** 1024×1024 | **Steps:** 24 | **CFG:** 3.6 | **Time:** ~**176.48s** ```text anime, masterpiece, best quality, 1girl, shrine maiden, long black hair, red hakama, white kimono top, holding paper talisman, sacred shrine background, cherry blossoms falling, soft sunlight, detailed face, serene expression, traditional japanese architecture, torii gate in background, cinematic lighting, depth of field ``` ![Test 1: Elegant Shrine Maiden](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta2aio_test1_shrine_maiden.jpg) **Test 2: Cyberpunk Street Scene** **Resolution:** 1536×1024 | **Steps:** 28 | **CFG:** 4.0 | **Time:** ~**229.52s** ```text anime, 2k quality, ultra-detailed, 1girl, cyberpunk hacker, neon-lit tokyo street, rain reflections, holographic advertisements, purple and cyan color scheme, tech wear jacket, mechanical arm augmentation, confident pose, sharp focus, cinematic composition, bokeh background, night city atmosphere, detailed eyes ``` ![Test 2: Cyberpunk Street Scene](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta2aio_test2_cyberpunk_street.jpg) **Test 3: Fantasy Dragon Knight** **Resolution:** 832×1216 | **Steps:** 32 | **CFG:** 3.8 | **Time:** ~**227.94s** ```text anime, masterpiece, high detail, 1girl, dragon knight, silver armor with blue accents, flowing cape, dragon companion beside her, epic fantasy landscape, castle ruins background, dramatic sky, wind effect on hair and cape, detailed armor patterns, heroic pose, cinematic lighting, depth of field ``` ![Test 3: Fantasy Dragon Knight](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta2aio_test3_dragon_knight.jpg) #### 🟡 Pruned Model fp8 (6-8 Steps Version) **Test 4: Cozy Cafe Moment** **Resolution:** 1024×1024 | **Steps:** 8 | **CFG:** 1.0 | **Time:** ~**32.47s** ```text anime, best quality, 1girl, casual outfit, sitting in cafe, holding coffee cup, warm lighting, bokeh background, soft smile, detailed eyes, cozy atmosphere, window light, autumn colors, relaxed pose ``` ![Test 4: Cozy Cafe Moment](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta2aio_test4_cozy_cafe.jpg) **Test 5: Magical Girl Transformation** **Resolution:** 512×768 | **Steps:** 7 | **CFG:** 1.0 | **Time:** ~**19.28s** ```text anime, masterpiece, 1girl, magical girl, transformation pose, sparkles and light effects, flowing hair, colorful costume, magic circle background, dynamic composition, vibrant colors, detailed ribbons, glowing effects ``` ![Test 5: Magical Girl Transformation](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta2aio_test5_magical_girl.jpg) **Test 6: Beach Sunset Portrait** **Resolution:** 1024×1536 | **Steps:** 6 | **CFG:** 1.0 | **Time:** ~**32.07s** ```text anime, best quality, 1girl, summer dress, beach sunset, golden hour lighting, ocean waves, soft wind effect on hair, warm colors, peaceful expression, detailed face, cinematic sunset, depth of field, romantic atmosphere ``` ![Test 6: Beach Sunset Portrait](https://huggingface.co/SeeSee21/QWEN-Anime/resolve/main/images/inline/beta2aio_test6_beach_sunset.jpg)

📝 Prompting tips

### ✍️ General Tips **Quality Tags (All Versions):** ```text anime, masterpiece, best quality, ultra-detailed, 2k resolution, sharp focus, cinematic lighting ``` **Style Modifiers:** ```text MOE STYLE, official art, anime coloring, detailed eyes, depth of field, bokeh ``` **Negative Prompt:** ```text low quality, blurry, bad anatomy, bad hands, text, watermark, mutation, distorted ``` ### 🎨 Beta3-AIO Specific Tips **Text-to-Image Prompts:** ```text anime girl with long blue hair, wearing school uniform, cherry blossoms in background, soft lighting, detailed eyes, anime style, high quality ``` **Image Editing Prompts (Single Image):** ```text change hair color to pink, add cat ears, school uniform, keep the same pose and composition ``` **Image Editing Prompts (Multiple Images):** ```text combine the character from image 1 with the background from image 2, match the lighting and style, anime aesthetic ``` **Important for Editing:** - Be specific about changes - Describe both images and desired result - Mention style consistency if needed - Natural language works best

🔧 Installation guide

### 📦 Beta3-AIO (Easiest!) 1. Download Beta3-AIO FP8 2. Place in `ComfyUI/models/checkpoints/` 3. Load with standard "Load Checkpoint" node 4. **For Image Editing:** Use provided workflow 5. Generate! **No extra files needed!** ### 📦 Beta2-AIO 1. Download your preferred version (Full or Pruned) 2. Place in `ComfyUI/models/checkpoints/` 3. Load with standard "Load Checkpoint" node 4. Generate! **No extra files needed!** ### 🧠 Beta2 (Safetensors) 1. Download checkpoint → `diffusion_models/` 2. Download Text Encoder → `text_encoders/QWEN/` 3. Download VAE → `vae/QWEN/` 4. Use "Load Diffusion Model" node ### 💾 Beta2 (GGUF) 1. Install ComfyUI-GGUF: [https://github.com/city96/ComfyUI-GGUF](https://github.com/city96/ComfyUI-GGUF) 2. Download GGUF → `unet/` 3. Download Text Encoder + VAE (same as Safetensors) 4. Use "GGUF Loader" node

🔧 Beta3-AIO specifications

**Base Model:** Qwen Image Edit 2511 · **Text Encoder:** Qwen 2.5 VL 7B FP8 (uncut) · **Precision:** FP8 · **Format:** AIO (All-in-One) · **File Size:** ~27 GB · **VRAM:** 8 GB minimum · **Steps:** 4-20 (4 min, 20 recommended) · **CFG:** 3.5 · **Sampler:** Euler · **Scheduler:** Beta **Capabilities:** - Text-to-Image generation - Image-to-Image editing (1-3 images) - Character combination - Scene composition - Style transfer

❓ FAQ

### General Questions **Q: Which version should I download?** A: Beta3-AIO for latest features + image editing. Beta2-AIO Pruned for fastest text-to-image. **Q: Do I need separate VAE/encoder for AIO versions?** A: No! AIO has everything integrated. **Q: Can I use Lightning LoRAs?** A: Yes! All versions support Lightning LoRAs (4-step or 8-step). ### Beta3-AIO Specific **Q: How many images can I edit at once?** A: 1-3 images (Image 1 required, Images 2-3 optional). **Q: Can I still do text-to-image with Beta3?** A: Yes! Beta3 supports both Text-to-Image AND Image-to-Image. **Q: What's the minimum steps for Beta3?** A: 4 steps minimum, but 20 steps recommended for quality. ### Beta2-AIO Specific **Q: What's the difference between Full and Pruned AIO?** A: Full = quality mode (CFG 2.5-4.0, 20+ steps). Pruned = speed mode (CFG 1.0, 6-8 steps). **Q: Why does Pruned need CFG 1.0?** A: It's optimized for low-step high-speed generation. CFG 1.0 works best. **Q: Can I use CFG 3.0 with Pruned?** A: Not recommended. Max is 2.0, but results are best at 1.0. ### Compatibility **Q: Is quality different between AIO and Beta2 FP8?** A: No, same training - AIO just bundles files together. **Q: Which has better quality: Full AIO or Beta2 BF16?** A: Beta2 BF16 has slightly better precision, but the difference is minimal. **Q: Can I mix versions (e.g., Beta3 with Beta2 VAE)?** A: Not recommended. Each version is optimized as a complete package.

🚀 Quick start (short version)

### Getting Started (Beta3-AIO) 1. **Download** Beta3-AIO FP8 from files section 2. **Place** in `ComfyUI/models/checkpoints/` 3. **Download** the provided workflow 4. **Load** workflow in ComfyUI 5. **Choose mode:** - Text-to-Image: Write prompt, generate - Image Editing: Upload 1-3 images, write edit prompt, generate 6. **Generate** amazing anime art!

ℹ️ Version information

**Current Version:** Beta3-AIO ⭐ · **Previous Versions:** Beta2-AIO, Beta2, Beta1 · **Release Date:** December 2025 · **License:** Apache 2.0 · **Format:** Safetensors (AIO)

🔞 Content notice

The original Civitai page lists limited nudity capabilities (no full explicit content) and marks the model for adults (18+) only. Use responsibly.

## 🙏 Credits **Training:** Custom dataset, Dual Tesla P40 GPUs · **Base Model:** Qwen Image Edit 2511 (Beta3), 2509 (Beta2) · **Text Encoder:** Qwen 2.5 VL 7B FP8 (uncut for Beta3) · **Architecture:** Qwen-Image-Edit framework · **Community:** Thanks to all Beta1, Beta2, and Beta3 testers! **Special Thanks:** - Qwen team for the base models - ComfyUI community for feedback - All users who provided testing data ## 📜 License - Base models: [Qwen/Qwen-Image-Edit-2511](https://huggingface.co/Qwen/Qwen-Image-Edit-2511) and [Qwen/Qwen-Image-Edit-2509](https://huggingface.co/Qwen/Qwen-Image-Edit-2509) (Apache-2.0); text encoder based on [Qwen/Qwen2.5-VL-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-VL-7B-Instruct) (Apache-2.0). This repository uses Apache-2.0 as stated on the original page. - Related: [Illustrious-Anime-Collection](https://huggingface.co/SeeSee21/Illustrious-Anime-Collection) (Qwen-Anime LoRAs). **Created with ❤️ for the anime AI community** Choose Beta3-AIO for the complete experience! - [SeeSee21 on Hugging Face](https://huggingface.co/SeeSee21) · [SeeSeeLP on Civitai](https://civitai.red/user/SeeSeeLP) ---

🧭 More from SeeSee — all models

### 🧩 Checkpoints & AIO | Model | Hugging Face | Civitai | |:--|:--|:--| | AniSee | [SeeSee21/AniSee](https://huggingface.co/SeeSee21/AniSee) | [Model](https://civitai.red/models/2628747) · [Workflows](https://civitai.red/models/2630318) | | Z-Anime | [SeeSee21/Z-Anime](https://huggingface.co/SeeSee21/Z-Anime) | [Model](https://civitai.red/models/2483351) | | Z-Image-Turbo-AIO | [SeeSee21/Z-Image-Turbo-AIO](https://huggingface.co/SeeSee21/Z-Image-Turbo-AIO) | [Model](https://civitai.red/models/2173571) · [Workflow](https://civitai.red/models/2174008) | | ERNIE-AIO | [SeeSee21/ERNIE-AIO](https://huggingface.co/SeeSee21/ERNIE-AIO) | [Model](https://civitai.red/models/2551548) · [Workflow](https://civitai.red/models/2553748) | | Chroma-Anime-AIO | [SeeSee21/Chroma-Anime-AIO](https://huggingface.co/SeeSee21/Chroma-Anime-AIO) | [Model](https://civitai.red/models/2022057) · [Workflow](https://civitai.red/models/2027641) | | FLUX.2-klein-4B-AIO | [SeeSee21/FLUX.2-klein-4B-AIO](https://huggingface.co/SeeSee21/FLUX.2-klein-4B-AIO) | [Model](https://civitai.red/models/2327389) · [Workflow](https://civitai.red/models/2327746) | | QWEN-Anime | [SeeSee21/QWEN-Anime](https://huggingface.co/SeeSee21/QWEN-Anime) | [Model](https://civitai.red/models/2122738) · [Workflow](https://civitai.red/models/2135240) | | SeeMoss-Genesis | [SeeSee21/SeeMoss-Genesis](https://huggingface.co/SeeSee21/SeeMoss-Genesis) | [Model](https://civitai.red/models/3031179) · [Workflow](https://civitai.red/models/3031981) | | See-Krea2 | [SeeSee21/Krea-2-See-Krea2-Genesis](https://huggingface.co/SeeSee21/Krea-2-See-Krea2-Genesis) | [Model](https://civitai.red/models/2846618) · [Workflow](https://civitai.red/models/2936142) | ### 🎀 LoRAs | Model | Hugging Face | Civitai | |:--|:--|:--| | Illustrious Anime Collection | [SeeSee21/Illustrious-Anime-Collection](https://huggingface.co/SeeSee21/Illustrious-Anime-Collection) | [Model](https://civitai.red/models/1994924) | | Chroma Semi-Anime | [SeeSee21/chroma-Semi-Anime](https://huggingface.co/SeeSee21/chroma-Semi-Anime) | [Model](https://civitai.red/models/1992501) | | Kitsune Furry Girl | [SeeSee21/Kitsune-Furry-Girl-Z-Image](https://huggingface.co/SeeSee21/Kitsune-Furry-Girl-Z-Image) | [Model](https://civitai.red/models/2218232) | | Ayase Momo (Dandadan) | [SeeSee21/Ayase-Momo-Z-Image](https://huggingface.co/SeeSee21/Ayase-Momo-Z-Image) | [Model](https://civitai.red/models/2189797) |

[🧩 Checkpoints & AIO](https://huggingface.co/collections/SeeSee21/checkpoints-and-aio) · [🎀 LoRAs](https://huggingface.co/collections/SeeSee21/loras) · [🌐 Civitai](https://civitai.red/user/SeeSeeLP) · [▶️ YouTube](https://www.youtube.com/@SeeSee-AI) _{Made with ❤️ by SeeSee}