Wan 2.7
Modelo de vídeo de código aberto com boa qualidade e versatilidade para cenas e animações.

Por que usar o Wan 2.7
Qualidade e versatilidade open source
Boa qualidade
Resultados sólidos em diversos tipos de cena.

Versátil
De cenas realistas a estilizadas, adapta-se ao prompt.

Comunidade ativa
Modelo open source com evolução constante.

Region Editing
Change only the part of the image you want to adjust. Select a region, describe the edit, and the rest of the frame stays untouched. Rework a background or swap a detail without regenerating everything.

Unified Generation and Editing
Create and refine in one continuous workflow. Start from a prompt, then keep editing the same image instead of starting over. A single model handles both directions, so nothing gets lost in between.

Fast Iteration
Produce multiple versions quickly to explore directions before you commit. Compare colorways, compositions, or styles side by side in seconds. Iterate until one option feels right, then polish it into the final asset.

Exemplos do Wan 2.7
Vídeos criados com o modelo
Write a prompt or upload a reference
Type a detailed description of the image you want, or upload a reference image to set the direction. Be specific about subject, composition, and mood to get a stronger first result.
Generate your first version
Pick Wan 2.7 Image in the Mitu workspace and run the generation. The model returns a full-resolution draft in seconds that you can build on.
Refine details with editing
Adjust colors, swap regions, or rework elements using the same image. Fine-tune until the output matches exactly what you had in mind.
Download and share your creation
Save your final image at full resolution and share it with your team or audience. Export it ready for web, social, and print use.
Como usar
Crie com o Wan 2.7 em três passos

Rainy evening street scene

Detailed pet portrait

Product-style food photography

Architectural interior rendering

Travel portrait photography

Dynamic animal capture
Perguntas frequentes
Explore outros modelos de IA
Modelos em destaque
