Wan 2.2 Image

Generate beautiful, cinematic 2MP images quickly.

Model Overview

This model is an optimized version of the Wan 2.2 model, designed to generate high-quality, cinematic images at a 2-megapixel resolution. It achieves this with impressive speed, producing results in just 3-4 seconds.

Best At

Creating stunning, cinematic-style images with a focus on aesthetic appeal and visual quality. It excels at generating images from detailed text prompts, offering control over aspect ratio and output quality.

Limitations / Not Good At

While powerful, it is primarily an image generation model. It does not handle video, audio, or complex reasoning tasks. Specific details like accurate text rendering within images might also be a challenge, as with most image generation models.

Ideal Use Cases

• Generating illustrations for blog posts or articles.
• Creating marketing visuals and social media content.
• Producing concept art or mood boards.
• Mockups for design projects.

Input & Output Format

• Input: Text prompt, optional parameters for "juiced" inference, aspect ratio, megapixels, seed, output format, output quality, and LoRA weights.
• Output: Image file (URI format).

Performance Notes

• Generates 2-megapixel images in 3-4 seconds.
• The juiced parameter can enable faster inference with additional optimizations.
• Supports LoRA weights for fine-tuning the generation style.