qwen-txt2img

1# Qwen Image 2512 Text-to-Image Workflows
2 
3## Overview
4 
5Qwen Image 2512 is the latest (December 2025) text-to-image model from the Qwen family. It uses a vision-language model (Qwen2.5-VL) as the text encoder and generates high-quality images from natural language prompts. Two workflow approaches:
6 
71. **QwenImageIntegratedKSampler**: All-in-one node (recommended for simplicity)
82. **Separate component loading**: UNETLoader + CLIPLoader + VAELoader + standard KSampler (more flexible)
9 
10## Models
11 
12### Standard Components
13 
14| Component | Node | Model | Notes |
15|-----------|------|-------|-------|
16| **UNET** | `UNETLoader` | `qwen_image_2512_fp8_e4m3fn.safetensors` | FP8, not currently installed — download if needed |
17| **CLIP** | `CLIPLoader` (type=`qwen_image`) | `qwen_2.5_vl_7b_fp8_scaled.safetensors` | Shared across all Qwen models, in clip/ |
18| **VAE** | `VAELoader` | `qwen_image_vae.safetensors` | Qwen-specific VAE (242MB) |
19 
20### Fine-tuned Variants (Installed)
21 
22| Model | Path | Focus |
23|-------|------|-------|
24| `qwenImageEditRemix_v10` | `diffusion_models/qwenImageEditRemix_v10.safetensors` | General-purpose remix |
25| `qwenUltimateRealism_v11` | UNETLoader path | Product photography, hyper-realistic |
26| `copaxTimeless` | UNETLoader path | Ultra-realistic portraits |
27| `qwnImageEdit_v16Bf16` | UNETLoader path | Abliterated (uncensored) |
28 
29## Lightning LoRAs
30 
31### 4-Step Lightning (General Qwen / txt2img)
32 
33```json
34{
35  "class_type": "LoraLoaderModelOnly",
36  "inputs": {
37…