WaveSpeed.ai
DashboardExploreInspiration
Home/Explore/wavespeed-ai/qwen-image/edit-plus-lora

image-to-image

wavespeed-ai/qwen-image/edit-plus-lora

Qwen-Image-Edit-Plus — a 20B MMDiT model for next-gen image edit generation with improved multi-image editing, single-image consistency, and native support for ControlNet.

Doc
preview
If enabled, the output will be encoded into a BASE64 string instead of a URL. This property is only available through the API.
If set to true, the function will wait for the image to be generated and uploaded before returning the response. It allows you to get the image directly in the response. This property is only available through the API.

Idle

The girl made a heart shape with her hands.

Your request will cost $0.025 per run.

For $1 you can run this model approximately 40 times.

One more thing:

ExamplesView all

The girl made a heart shape with her hands.
Based on the woman in Figure 2 and the man in Figure 1, generate a wedding photo set, following these descriptions: The groom wears a red Chinese-style tunic, and the bride wears an exquisite Xiuhe dress, with a golden phoenix coronet on her head. They stand side by side in front of an ancient vermilion palace wall, with carved wooden windows in the background. The lighting is bright and soft, the composition is symmetrical, and the atmosphere is festive and solemn.
The girl in Figure 1 is sitting in the studio in Figure 2, speaking into the microphone.
Change to a blue ID photo, with the person in the picture wearing a white-collar shirt with a silk texture.
The girl in Figure 1 sits in the pose of Figure 3 wearing the black dress from Figure 2.
Put this air conditioner in the living room next to the sofa.

README

Qwen-Image-Edit-Plus (20B, MMDiT)

A next-gen image editing model built on Qwen-Image 20B. It delivers precise bilingual (Chinese & English) text editing, supports both appearance-level and semantic-level edits, and preserves the original style.

What choose this?

  • Dual-mode editing

    • Appearance editing: add/remove/modify elements while keeping all other regions pixel-accurate and unchanged.
    • Semantic editing: higher-level changes—IP creation, pose/rotation, style transfer—allow global pixel updates while keeping semantic intent.

  • Precise text editing (CN/EN) Edit on-image text directly (add/delete/replace) while retaining the original font, size, kerning, and style.

  • Style preservation Maintains palette, lighting, brushwork, and overall look even under substantial edits.

  • Strong benchmark results Evaluated across multiple public editing benchmarks with state-of-the-art performance.

Designed for

  • Design & Marketing teams – Rapid visual iterations, brand-safe edits, and multilingual comps.
  • E-commerce & Social – Clean product touch-ups, quick hero swaps, localized text.
  • Creators & Studios – Concepting, IP style moves, pose/angle changes without repainting.

Example prompts

  • Appearance (CN): 在桌面右上角添加一杯拿铁，不改变其他区域。
  • Semantic (EN): Turn the product into a cyberpunk style while keeping the brand logo and layout consistent.
  • Text edit (EN): Replace the headline "Summer Sale" with "Autumn Sale" and keep the same font and size.

Pricing

Just $0.02 per image !!!

How to use

  1. Upload the source image.
  2. Write the prompt (Chinese or English).
  3. Generate — results arrive in moments.
  4. Review & iterate — keep the same seed for exact reproduction, or change it for A/B comparisons.