Wan 2.5 Image to Image API

Deploy the enterprise-grade Wan 2.5 image to image API for semantic restyling, relighting, and identity-locked editing. Supports multi-reference inputs and high-fidelity output via EvoLink.

Configure Request

Response

Click "Generate" to see results

Wan 2.5 Image to Image API: Consistent Semantic Restyling

Leverage the Wan 2.5 image to image API to transform visuals while locking face identity and layout. Ideal for product photography, anime adaptation, and professional creative workflows.

Capabilities of the Wan 2.5 Image to Image API

Identity-Locked Style Transfer

Use the Wan 2.5 image to image API to apply cinematic, anime, or oil painting styles while strictly preserving facial features and body pose.

Explore Styles

Commercial Product Restyling

Automate e-commerce workflows by swapping backgrounds and lighting using the Wan 2.5 image to image API without distorting product details.

See Showcase

Instruction-Based Inpainting

Perform localized edits with brush masks. The Wan 2.5 image to image API allows for removing text or changing props seamlessly.

View Documentation

Why Developers Choose the Wan 2.5 I2I API

Engineering teams prefer the Wan 2.5 image to image API for its balance of creative control, cost-efficiency, and reliable uptime via EvoLink.

Superior Consistency

Unlike standard diffusion models, Wan 2.5 maintains structural integrity, making it safe for brand assets.

Scalable Token Pricing

The Wan 2.5 image to image API offers tiered token-based pricing, optimizing costs for high-volume batch processing.

Granular Editing Control

Gain pixel-level influence over outputs with advanced masking and prompt guidance features.

How to Integrate the Wan 2.5 Image to Image API

Three simple steps to automate high-quality image restyling.

Step 1 — Send Reference & Mask

POST your base image to the Wan 2.5 image to image API endpoint. Optionally includes up to 5 reference frames or inclusion masks.

Step 2 — Define Style Parameters

Configure the JSON payload with style prompts, negative prompts, and `denoising_strength` to control the transformation intensity.

Step 3 — Retrieve & Refine

Receive 1024–1536px generated assets. Use the API to iteratively refine results based on feedback loops.

Open API Playground

Wan 2.5 API Key Features

Technical specifications for the Wan 2.5 image to image API

Consistency

Robust Identity Preservation

Market-leading algorithm for keeping faces/objects consistent.

Editing

Smart Masking Support

Native API support for alpha masks and inpainting regions.

Control

Precision Style Control

Fine-tune aesthetic shifts with floating-point strength values.

Lighting

Advanced Relighting

Natural language prompts to alter scene lighting and mood.

Quality

High-Res 1536px Output

Generate sharp, print-ready visuals directly from the API.

Cost

Efficient Token Economy

Pay only for the compute you use with the Wan 2.5 image to image API.

Wan 2.5 I2I vs. Competitors

Compare the Wan 2.5 image to image API against other models

Model	Duration	Resolution	Price	Strength
Wan 2.5 Image-to-Image	N/A	1024–1536px	Low (Token-based)	Best-in-class identity lock, API-first design, precise masking.
Wan 2.5 Text-to-Image	N/A	1024–1536px	Token-based	Higher creativity freedom but lower structural consistency.
Standard Diffusion XL	N/A	1024px Base	Varies by provider	General purpose, but often struggles with strict identity retention.

Wan 2.5 Image to Image API FAQs

Everything you need to know about the product and billing.

The Wan 2.5 image to image API is a developer interface for accessing the WanXiang 2.5 model, allowing for high-fidelity image restyling, inpainting, and style transfer while preserving the original image's structure and identity.

While T2I creates images from scratch, the Wan 2.5 image to image API uses an existing image as a structural blueprint. This ensures the output retains the exact composition, pose, and facial features of your reference.

Yes, the Wan 2.5 image to image API supports multi-reference conditioning (up to 5 images), which helps the model better understand complex subjects or specific 360-degree product details.

The API natively supports high-resolution outputs ranging from 1024px to 1536px on the longest edge, suitable for professional creative usage.

Absolutely. EvoLink provides a commercial-grade SLA for the Wan 2.5 image to image API, ensuring data privacy, high availability, and clear usage rights for generated assets.

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 reference images. Supports image-to-image generation mode.

Asynchronous processing mode, use the returned task ID to .

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

Request Parameters

modelstringRequiredDefault: wan2.5-image-to-image

Image generation model name.

Examplewan2.5-image-to-image

promptstringRequired

Prompt describing how to transform the input image.

Notes

Limited to 2000 tokens

ExampleTransform this image into a watercolor painting

nintegerOptionalDefault: 1

Number of images to generate.

Value	Description
1-4	Batch generation supported

Notes

Pre-charged based on n value, actual charge based on generated count

Example1

image_urlsarrayRequired

Reference image URL list for image-to-image transformation.

Notes

Max 2 images per request
Max size: 10MB per image
Formats: .jpeg, .jpg, .png (no transparent), .bmp, .webp
Resolution: width and height range [384, 5000] pixels
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": "wan2.5-image-to-image",
  "prompt": "Transform this image into a watercolor painting",
  "image_urls": [
    "https://example.com/image1.png"
  ]
}

Response Example

{
  "created": 1757165031,
  "id": "task-unified-1757165031-uyujaw3d",
  "model": "wan2.5-image-to-image",
  "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": 1.2,
    "user_group": "default"
  }
}