A Qwen Image 2.1 studio workflow. One model load, then four jobs you turn on one at a time: text to image, single-image edit, multi-image edit, and a pose editor. Each job can upscale with SeedVR2 and save a PNG with Civitai metadata.<hr />## What you can do- Text to image — prompt in, image out, then an optional SeedVR2 upscale.- Single image edit — change a photo from a prompt. A mask is optional. Use one when the area is small or when several similar objects are in the frame. Put the photo on image 1 and the mask on image 2, and say <code>&lt;image1&gt;</code> and <code>&lt;image2&gt;</code> in the prompt (no space). Tell it to change only the masked region, or the edit spreads over the whole picture.- Multi-image edit — several reference images in one instruction.- Pose editor — VNCCS Pose Studio sets the pose, your photo sets the person. Image 1 is the pose render. Image 2 is the character photo. The positive prompt is <code>Draw character from image2</code>. Pose Studio’s view size and the sampler size stay the same (1024×1024). Keep the mannequin’s muscle, weight, and height near the middle.Use the group bypasser and leave only one job enabled.<hr />## Settings- Steps 8, CFG 1, denoise 1, with the 8-step LoRA on. That LoRA is built for CFG 1. Higher CFG fights the pose and breaks hands, arms, and legs.- Generate at 2MP or higher. Use the Resolution Master presets. The Krea 2 Turbo set is a good default. The name is only a label.- Pose LoRA <code>VNCCS_QI2_PoseStudioV1.1</code> stays at strength 1, in front of the 8-step LoRA.- SeedVR2 tile size stays at 1024. Tile count 2 is a normal 2× upscale. Higher counts run one full pass per tile. On a 16 GB GPU, use the 3B SeedVR2 model and leave Cache Model on. Keep Manual Load bypassed unless you mean to use it.<hr />## SavingEach job has a normal save and a Civitai save. The Civitai save writes the prompt, seed, steps, CFG, sampler, scheduler, and model name into the PNG. Sampler settings for each job live on that job’s sampler settings node.<hr />## You will need- UNET: <code>qwen_image_2.1_int8_convrot</code>- VAE: <code>qwen_image_2.1_vae_bf16</code>- Text encoders used in the graph: <code>qwen3vl_8b_nvfp4_heretic</code>, <code>qwen3.5_9b_qwen_image_2.1_pe_t2i</code>, <code>qwen3.5_9b_qwen_image_2.1_pe_i2i</code>- LoRAs: <code>p_qwen_image_2.1_8step_v0.1</code>, <code>VNCCS_QI2_PoseStudioV1.1</code>- A SeedVR2 upscale model if you use that stageTested on an RTX 5070 Ti 16 GB.

Qwen Image 2.1 Studio | T2I | I2I | REF2I | POSE | Beginner Friendly

Version 1.0 Model- This version uses Qwen Image 2.1 Turbo.- The smaller int8 turbo file is the one to start with. The larger bf16 turbo is there if you want higher quality and have the memory for it.- The old model file and the 8-step LoRA are no longer part of the workflow. How you use it- One Mode menu: new image, edit one photo, edit with several photos, or pose.- One prompt box for every mode. You can preview the prompt that was actually used.- Pose: make the pose, load the person, and queue. The pose LoRA and the pose prompt turn on only for that mode.- For edits, you can keep the photo's own size, or turn on the switch that uses the resolution you picked.- Several photos: the result follows the size of photo 1. Empty slots are ignored. You can add more than 6 if you need them, up to 16.- LoRAs are still optional, split into style, character, and edit. Upscale and finish- SeedVR2 can be off, on after the generation, or used on your own picture with no generation.- Size is a simple multiply. 2 means twice as wide and twice as tall.- You can mix SeedVR2 with a plain upscale, so skin stays softer.- Tiles are lined up better, so the seams are less likely to show.- Film grain is optional and always last. Save- One save for every mode, instead of a separate save group for each.- Optional helper so the saved image carries the info Civitai expects. Clean-up- Groups went from 34 down to 13. Nodes on the canvas went from 92 down to 56.- Removed the optional speedup nodes. Those were for the old model.- Removed the nodes used to bypass groups, pass images around, show the prompt, time the run, and switch between saves.

Files with SHA256: 27512b0bc7f303e2115d37fe10e7743a0763b458f6656838d75446ce08948264