Wan 2.7 Image API

Text to ImageImage to Image

Fast AI image generation with thinking mode and editing support. Max 2K resolution.

·Features·FAQ

Documentation

Prompt *

Text description of the image (max 5000 chars)

Aspect Ratio

Reference Images

Input images for editing (up to 9 images, 100MB each)

0/9

Click or drag & dropPNG, JPG, WEBP, GIF · Max 100MB

Thinking Mode

Better quality, slower. Text-to-image only

Seed

Seed for reproducibility

Output

Your generated image will appear here

Features

What Wan 2.7 Image API offers

Faster generation than Wan 2.7 Image Pro in the same family

Output up to 2K resolution

Thinking mode for text-to-image when no reference images are provided

Up to nine reference images for image-guided generation

Seven aspect ratio presets

Optional seed for reproducibility

Use cases

Built for

Primary

Production pipelines: faster turnaround while staying on Wan 2.7 image

Web and social: 2K assets for screens and responsive layouts

Multi-reference creative: combine several images for style or subject alignment

Iteration: thinking mode off or on depending on prompt difficulty

FAQ

About Wan 2.7 Image API

This variant is tuned for faster runs and 2K-class output. Pro targets higher peak resolution (up to 4K). Both accept up to nine reference images and share the same aspect ratio list. Thinking mode applies to plain text-to-image runs without references (see the playground tooltip).

Yes. You can attach up to nine images. When references are present, the request follows the editing and image-guided path described in the docs instead of applying thinking mode.

When you generate from text only, thinking mode spends extra reasoning before rendering for better coherence. It can increase latency. It is not applied when reference images are supplied.