Qwen Pose/Position Edit (instruction)

Image → Image

Edit an existing image from a plain-English instruction — change a subject's pose, position, or what they're doing while holding their identity, clothing, the room, and lighting. Powered by Qwen-Image-Edit-2511; the source image is the reference (no LoRA).

qwen_pose_edit12 GB VRAM30.2 GB of weights

Calling it

FlixML has no generate button: you tell your agent what you want and it calls the API on your install. The id is what names this workflow.

POST /api/image/generate
{
  "workflow": "qwen_pose_edit",
  "image": 0,
  "prompt": 0
}

Parameters

imagestringrequired

Source image filename or Studio output path to edit

promptstringrequired

Plain-English edit instruction. Describe only what changes; the model keeps everything else. Name each body part and where it goes ('left hand on her left hip'). Place the subject against named objects with an exact position ('on the path, one step left of the bench, not touching it'). 'Next to the bench' is too loose and can put the subject in front of the bench instead.

seedintegerdefault 42

Your install is the source of truth: GET /api/workflows returns these parameters as they are on your machine.

What it costs to run

TESTED 2026-09-14 on RTX 4070 Ti (12 GB) + 16 GB system RAM, 20 steps at 832x1248, same prompt and seed: the fp8 file (20.5 GB) sampled at 5.8 s per step, the Q4_K_M .gguf (13.2 GB) at 9.7 s per step, with near-identical output. The fp8 file is larger but the card does fp8 math natively and ComfyUI's Dynamic VRAM streams it from disk. A Q3_K_L .gguf (10.6 GB) was no faster than Q4_K_M and drew worse anatomy.

Same input, different job

FLUX.2 Klein Reference Edit

Generate a new image from a reference image and a plain-English instruction. FLUX.2 Klein 4B reads the reference as conditioning and samples a fresh frame, so unlike img2img it can change pose, wardrobe and setting while holding the subject's identity. Say what changes and what stays.

Qwen Multi-Angle (re-angle a still)

Re-shoot an existing image of the same subject from a new camera angle. Qwen-Image-Edit-2511 + multi-angle LoRA holds identity, clothing, and lighting while changing the camera viewpoint. Produces a sibling of the source still at a different angle.

SDXL Image-to-Image

Generate SDXL image variations from a source image using prompt guidance and denoise strength