Alibaba WAN 2.2 Text-to-Video Model (1080p)

Alibaba WAN 2.2 is an advanced text-to-video (T2V) model powered by an MoE (Mixture of Experts) architecture. It generates cinematic-quality videos at Full HD 1080p resolution, supporting both landscape (1920×1080) and portrait (1080×1920) formats.

Why it looks great

• Cinematic quality: controls lighting, color, and camera composition for professional results.
• Smooth motion: handles subject and camera movement with stability and naturalness.
• Semantic alignment: faithfully follows detailed text prompts, even in complex scenes.
• Prompt expansion (optional): refine prompts automatically for enhanced output.

Limits and Performance

• Input: text prompt
• Output resolution: 1080p (1920×1080 or 1080×1920)
• Clip length per job: 5 seconds

Pricing

Duration	Resolution	Cost per job
5 s	1080p	$0.80

How to Use

1. Write Prompt – describe the scene, mood, motion, and camera style.
2. Choose Size – landscape (1920×1080) or portrait (1080×1920).
3. (Optional) Add a Negative Prompt to exclude unwanted details.
4. (Optional) Set Seed for reproducibility.
5. Run – preview and download your 5-second video.

Pro tips

• Use clear motion cues in the prompt (e.g., “slow pan”, “gentle breeze”).
• Choose portrait mode (1080×1920) for mobile/social content, landscape for cinematic use.
• Apply negative prompts to avoid artifacts like text, watermarks, or distortions.
• Enable prompt expansion if you want the system to refine under-specified prompts.

Notes

• Please check that your prompt and parameters are correct before running.
• If results don’t align, try re-running with different seeds.