Qwen-Image-Ed-2511-Hyper-Realistic-Portra

About

What Qwen-Image-Ed-2511-Hyper-Realistic-Portra is

Qwen-Image-Edit-2511-Hyper-Realistic-Portrait Qwen-Image-Edit-2511-Hyper-Realistic-Portrait is an adapter LoRA developed for Qwen’s Qwen-Image-Edit-2511 image-to-image model, specifically designed to transform images into ultra-realistic, photorealistic face portraits with strict identity preservation. The model enhances pore-level skin textures, realistic moisture effects, and natural wet hair clumping while preserving original facial structure and details. It applies physically plausible lighting, natural eye reflections, and shallow depth of field to achieve a true photographic look without artificial smoothing or retouching. Strong visual consistency is maintained across lighting, shadows, color tones, and composition, ensuring results remain authentic and camera-realistic. This makes the model well suited for high-fidelity portrait generation where realism and identity accuracy are critical. Above are example outputs demonstrating the model’s ability to deliver high-quality, visually cohesive hyper-realistic portrait results. [!important] Note: The face identity in the input image used for inference should be well defined to accurately reproduce a Hyper-Realistic Portrait (HRP). Quick Start with Diffusers Compatible with versions 2509 and 2511. Trained on ModelScope.cn Recommended: HRP20.safetensors Core Prompt Trigger Prompt Use the following prompt to activate the hyper-realistic portrait effect: Download Model You can download the model files from the Files & versions tab: Download reusable JSON reusable PROMPT Transform the image into an ultra-realistic photorealistic portrait with strict identity preservation, facing straight to the camera. Enhance pore-level skin textures, realistic moisture effects, and natural wet hair clumping against the skin. Apply cool-toned soft-box lighting with subtle highlights and shadows, maintain realistic green-hazel eye catchlights without synthetic gloss, and preserve soft natural lip texture. Use shallow depth of field with a clean bokeh background, an 85mm macro photographic look, and raw photo grading without retouching to maintain realism and original details.

Summarised from the published model card. Read the full card on the HuggingFace links below.

Specifications

What it is

How it works

How image models work

Variants

Sizes & precisions

Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.

Using it via the API

Call it like any OpenAI endpoint

Once AxForge deploys qwen-image-ed-2511-hyper-realistic-portra for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (qwen-image-ed-2511-hyper-realistic-portra below is illustrative; you get the exact model name on deployment.)

$ curl -sS https://api.axforge.ai/v1/images/generations \
  -H "Authorization: Bearer $AXFORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"qwen-image-ed-2511-hyper-realistic-portra","prompt":"a red bicycle","size":"1024x1024"}'

Details

Languages, data & research

Licence

Open weights

Sources

Weights & code

Explore

More image models