Image Generation ModelDeprecated

Wan 2.2

Wan 2.2 is an image generation model from Wan, available via Wavespeed, supporting LoRA fine-tuning and configurable resolution.

Start Building with Wan 2.2 View All Models

Publisher

Wan

TypeImage

Context Window10,000 tokens

Replaced byWan 2.7

Provider

WaveSpeed

LoRA

Try Wan 2.2 →

About Wan 2.2

Image generation with LoRA support

Wan 2.2 is an image generation model published by Wan and served through the Wavespeed provider on MindStudio. It accepts text prompts alongside optional image inputs, resolution selection, duration settings, and LoRA adapters, giving users direct control over the generation parameters. The model supports a context window of up to 10,000 tokens and is tagged for LoRA compatibility, meaning custom fine-tuned style or subject adapters can be applied at inference time.

Wan 2.2 is suited for workflows that require flexible image generation with the ability to steer outputs through LoRA weights, such as consistent character rendering, style transfer, or domain-specific visual content. Its selectable resolution and seed inputs make it reproducible and adaptable to different output size requirements. Note that this model is currently listed as deprecated on MindStudio, so users should verify availability before building production workflows around it.

Capabilities

What Wan 2.2 supports

LoRA Support

Accepts custom LoRA adapter weights at inference time, allowing fine-tuned styles or subjects to be applied directly to image generation without retraining the base model.

Text-to-Image Generation

Generates images from text prompt inputs, with a context window of up to 10,000 tokens to accommodate detailed descriptions.

Image Input

Accepts an image as an additional input alongside text, enabling image-conditioned generation or reference-guided outputs.

Resolution Selection

Provides a selectable resolution parameter so users can configure output image dimensions to match their specific use case.

Seed Control

Accepts a numeric seed value to make generation reproducible, allowing the same prompt and settings to produce consistent outputs across runs.

Ready to build with Wan 2.2?

Get Started Free

FAQ

Common questions about Wan 2.2

What is the context window for Wan 2.2?

Wan 2.2 has a context window of 10,000 tokens, which applies to the text prompt input used to guide image generation.

Does Wan 2.2 support LoRA adapters?

Yes. Wan 2.2 includes a dedicated LoRA input, allowing users to supply custom LoRA weights at inference time to apply fine-tuned styles or subjects.

What is the pricing for Wan 2.2 on MindStudio?

Pricing information for Wan 2.2 is not published in the available metadata. You should check MindStudio's current pricing page or the model's listing for up-to-date cost details.

Is Wan 2.2 still available for use?

Wan 2.2 is currently listed as deprecated on MindStudio. It may still be accessible but is no longer actively maintained on the platform, so users should confirm availability before building workflows that depend on it.

Can I control the output resolution and get reproducible results?

Yes. Wan 2.2 exposes both a resolution selector and a numeric seed input, so you can configure the output dimensions and lock in a seed value to reproduce the same result across multiple runs.

Resources