Qwen Image Edit API

Leverage Qwen Image Edit, Alibaba's advanced image editing API, for programmatic photo manipulation. Upload images, define masks, and execute complex instruction-based edits—from text correction to object removal—at scale.

Image Upload Required

Upload 1-3 reference images to start editing. Supported formats: jpg, png, webp, bmp, tiff.

Prompt

A cute fluffy cat captain sitting in a spaceship commander chair, wearing a small headset and a futuristic collar. Pixar animation style, 3D render, vibrant colors, soft volumetric lighting. The cockpit is cozy with round shapes, galaxy background outside the window. Octane render, high fidelity, adorable expression.

Parameters

image_urls

https://pub-ebac96ac8a0c4c789f53c7470fd2339c.r2.dev/BQACAgUAAyEGAASHRsPbAAEDFHJo3RjyLCNIDyRJPhulBDRTS2e1dQACEBsAAs7-6FbennQ2yXsmjzYE (4).jpg

Estimated Cost (per edit)

3 Credits

Sample Result

Prompt

318 (suggested: 2,000)

Input Images *

Upload up to 3 images (max 10MB each)

Click to upload or drag and drop

Supported formats: JPG, JPEG, PNG, BMP, WEBP, TIFF
Maximum file size: 10MB; Maximum files: 3

History

Max 20 items

0 running · 0 completed

Your generation history will appear here

Pricing

Starting from

$0.050

3 Credits

per edit

Guaranteed

99.9% Uptime

Automate image editing with Qwen Image Edit API

Execute precise, region-based image edits using natural language instructions. Qwen Image Edit supports mask-guided inpainting, bilingual text replacement, and background/object edits for high-volume production workflows.

Capabilities of the Qwen Image Edit API

Smart Object & Text Removal

Use Qwen Image Edit to remove unwanted elements (watermarks, logos, objects) and reconstruct the background with consistent lighting and texture.

Start Using API

Precise Text Replacement

A key strength of Qwen Image Edit: replace text on signs or posters while better preserving the original font style and visual consistency.

Start Using API

Batch Image Processing

Send up to 3 images per request to apply consistent retouching rules across a product set. Limits can vary by plan and configuration.

Start Using API

Why developers choose Qwen Image Edit API

Use Qwen Image Edit for instruction-following image edits with strong mask control and reliable text replacement for production workflows.

Instruction-Based Control

Use natural language (e.g., "change the shirt to red") to guide the Qwen Image Edit API without complex parameter tuning.

High Fidelity Inpainting

The model understands context, ensuring edited regions blend seamlessly with the original lighting and perspective.

Cost-Effective Scaling

Control costs with per-edit billing and predictable usage tracking. Effective cost can vary with input size and routing—verify in your dashboard.

How to integrate the Qwen Image Edit API

A simple request-response flow for complex image manipulations.

Step 1 — Construct the Payload

Send a POST request containing your base image (Base64/URL) and an optional mask image to the endpoint.

Step 2 — Define the Instruction

Pass a text prompt to the Qwen Image Edit API describing the desired change (e.g., "replace the cat with a dog").

Step 3 — Retrieve & Deploy

Receive the processed image URL in the JSON response, ready for immediate display in your application.

Open API Playground

API Key Features

Engineered for granular control and automation

Precision

Mask-Guided Inpainting

Restrict edits to specific pixels using binary masks.

Efficiency

Multi-Image Batching

Process small batches (up to 3 images) in a single API call (plan dependent).

Composition

Scene Restaging

Swap backgrounds while keeping subject depth correct.

Typography

Text Style Preservation

Modify embedded text without breaking visual design.

Global

Bilingual Instruction

Native support for complex prompts in English & Chinese.

Economy

Transparent Pricing

Transparent per-edit billing with dashboard-based usage tracking (pricing varies by routing and workload).

Qwen Image Edit API vs. Alternatives

Choose the right programmatic editor for your stack

Model	Category	Core capability	Output sizing	Billing	Best for
Qwen Image Edit API	Image editing	Mask-guided inpainting, text replacement, local retouching	Flexible (input-dependent)	Per successful edit (usage-based)	Programmatic image editing workflows where precise local edits and text replacement matter
Wan 2.5 Image-to-Image	Image variation / restyling	Reference-based variation and heavy restyling	High-res outputs (config dependent)	Usage-based	Creating variations and stylistic transformations from a reference image (typically less suited for precise local text edits)
Nano Banana 2	Text-to-image generation	General-purpose image generation	1024px base (config dependent)	Usage-based	General generation when you don't need mask-guided local editing or text-preserving retouching

Qwen Image Edit API FAQs

Everything you need to know about the product and billing.

The API supports concurrent requests, but standard tiers typically handle up to 3 images per batch call. Rate limits depend on your EvoLink plan.

Yes. Qwen Image Edit supports text replacement within images and is designed to better preserve style and layout. Results can vary by input image quality and typography complexity.

Masks should be binary images (black/white) where white represents the area to edit. They must match the resolution of the input image.

Yes, you can extend the canvas boundaries. Simply provide a mask covering the empty space and prompt the model to fill the new area.

The API accepts standard web formats (JPG, PNG) via URL or Base64 encoding, with a recommended file size under 10MB for optimal latency.

Pricing is usage-based and billed per successful edit. Effective cost can vary with input size and routing; use your dashboard usage and billing records as the source of truth.

API Reference

Select endpoint

Authentication

All APIs require Bearer Token authentication.

Header

Authorization: 
Bearer YOUR_API_KEY

Get API Key

POST

/v1/images/generations

Generate Image

Create an image generation task using text prompts or reference images. Supports text-to-image, image-to-image, and image editing modes.

Asynchronous processing mode, use the returned task ID to .

Generated image links are valid for 24 hours, please save them promptly.

Request Parameters

modelstringRequiredDefault: qwen-image-edit

Image generation model name.

Exampleqwen-image-edit

promptstringRequired

Prompt describing the image to be generated or how to edit the input image.

Notes

Limited to 2000 tokens

ExampleA cat playing in the grass

image_urlsarrayRequired

Reference image URL list for image editing.

Notes

Max 3 images per request
Image dimensions must be within 384-3072 pixels
Formats: .jpg, .jpeg, .png, .bmp, .webp, .tiff
URLs must be directly accessible by the server

Examplehttps://example.com/image1.png

callback_urlstringOptional

HTTPS callback address after task completion.

Notes

Triggered on completion, failure, or cancellation
Sent after billing confirmation
HTTPS only, no internal IPs
Max length: 2048 chars
Timeout: 10s, Max 3 retries

Examplehttps://your-domain.com/webhooks/image-task-completed

Request Example

{
  "model": "qwen-image-edit",
  "prompt": "Replace the background of this image",
  "image_urls": ["https://example.com/image1.png"]
}

Response Example

{
  "created": 1757165031,
  "id": "task-unified-1757165031-uyujaw3d",
  "model": "qwen-image-edit",
  "object": "image.generation.task",
  "progress": 0,
  "status": "pending",
  "task_info": {
    "can_cancel": true,
    "estimated_time": 45
  },
  "type": "image",
  "usage": {
    "billing_rule": "per_call",
    "credits_reserved": 1,
    "user_group": "default"
  }
}