Wan 2.2 Vace Inpainting overview
Wan 2.2 Vace Inpainting by Alibaba Cloud
Wan 2.2 VACE Inpainting API enables precise AI-powered image and video inpainting for controlled creative editing.
Wan 2.2 Vace Inpainting specs
- Model ID
wan-22-vace-fun-a14b/inpainting- Developer
- Alibaba Cloud
- Input
- Text, Video
- Output
- Video
- Endpoint
POST https://api.aimlapi.com/v2/video/generations
Overview
Wan 2.2 Vace Inpainting is an advanced video-to-video generative AI model designed primarily for high-quality inpainting tasks in video content. It facilitates seamless video editing by enabling users to mask and modify specific regions in videos, maintaining contextual continuity, motion consistency, and fine detail restoration. This model integrates cutting-edge multimodal understanding and adaptive video generation technologies, optimized for resolutions up to 720p.
Technical Specifications
- Model Architecture: Multimodal video and image transformer backbone with adaptive scene and motion prediction.
- Parameter Size: 14 billion parameters for fine granularity in video detail synthesis.
- Resolution Range: Up to 720p resolution output; default usage often at 480p for speed.
- Frame Rate Processing: Operates typically at 16 frames per second for stable video synthesis.
- Input/Output Formats: Supports mp4, mov, webm, m4v, gif for video; jpg, jpeg, png, webp, gif, avif for images.
- Memory Use: Designed for efficient GPU usage supporting local and cloud-based workflows with moderate memory footprint.
Performance Benchmarks
- Inpainting Quality: Excels in preserving context and texture detail during localized video edits.
- Temporal Stability: Strong performance in motion continuity compared to baseline video inpainting models.
- Resolution Scaling: High fidelity at 720p with minor trade-offs in processing speed; faster processing at 480p with maintained visual consistency.
- Maximum Clip Length: Typically supports up to 80-81 frames reliably, beyond which quality may degrade.
Key Features
- Video Inpainting with Masking: Allows selective editing by inputting a source video and a corresponding mask video to inpaint or replace masked areas.
- Temporal Consistency: Maintains smooth motion flow and coherence across frames to prevent flickering or artifacts during inpainting.
- High Detail Restoration: Reconstructs fine textures and details within the masked region for natural appearance.
- Resolution Support: Supports video output commonly at 480p, 580p, and up to 720p with quality scaling depending on resource availability.
- Flexible Input Types: Accepts various video formats including mp4, mov, webm, m4v, and gif, and image formats like jpg, png, webp for mask or reference inputs.
- Integration with ComfyUI: Compatible with ComfyUI workflows to combine inpainting with text-to-video, image animation, and outpainting pipelines.
Comparison with Other Models
vs Seedance 1.0: Seedance 1.0 provides slightly better frame rate output at 24 fps compared to Wan 2.2's 16 fps, beneficial for ultra-smooth video generation. However, Wan 2.2 offers better integration for inpainting flexibility and runs efficiently on more modest hardware setups, making it more accessible for creators.
vs Veo 3: Veo 3 leads on advanced closed-source video enhancement with superior resolution and speed but at a significantly higher cost. Wan 2.2 competes well as an open-source alternative with strong multimodal inpainting capabilities and easier API integration, suitable for broad professional use.
vs Generic Baseline Video Inpainting: Compared to basic baseline models, Wan 2.2 delivers significantly improved texture detail restoration and motion coherence. Baseline models often produce more artifacts and flickering, making Wan 2.2 the preferable choice for quality video inpainting projects.
API Integration
Accessible via AI/ML API. Documentation: available here.
Use cases
- Professional video post-production: VFX touch-ups, object removal, scene re-editing.
- Digital marketing content creation: Automated video personalization and brand adaptation.
- Educational video material enhancement: Visual reconstructions or content update animations.
- Creative arts and digital storytelling: Seamless animation replacements and effects.
Wan 2.2 Vace Inpainting API pricing
| Usage | Price | Unit |
|---|---|---|
| Video generation | $0.05 | per generation |
| Additional | $0.013 | per generation |
Prices from the AI/ML API price list, checked on Nov 4, 2026.
Wan 2.2 Vace Inpainting vs other video models: prices
AI/ML API rates for 5 models, per generation.
| Model | Per generation |
|---|---|
 Wan 2.2 Vace Inpainting | $0.05 |
| Wan 2.2 Vace Depth | $0.05 |
| Wan 2.2 Vace Outpainting | $0.05 |
| Wan 2.2 Vace Pose | $0.05 |
| Wan 2.2 Vace Reframe | $0.05 |
How to use the Wan 2.2 Vace Inpainting API
View docs
Build with your agent
Give your agent the model details, code and documentation.
1Get your API key
Create API keyexport AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl 'https://api.aimlapi.com/v2/video/generations' \
-H "Authorization: Bearer $AIMLAPI_KEY" \
-H 'Content-Type: application/json' \
-d '{
"model": "wan-22-vace-fun-a14b/inpainting",
"prompt": "A slow dolly shot through a sunlit greenhouse full of ferns, light dust in the air, natural camera motion",
"video_url": "https://example.com/input.mp3"
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "wan-22-vace-fun-a14b/inpainting",
"prompt": "A slow dolly shot through a sunlit greenhouse full of ferns, light dust in the air, natural camera motion",
"video_url": "https://example.com/input.mp3"
}
response = requests.post("https://api.aimlapi.com/v2/video/generations", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = await fetch("https://api.aimlapi.com/v2/video/generations", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "wan-22-vace-fun-a14b/inpainting",
"prompt": "A slow dolly shot through a sunlit greenhouse full of ferns, light dust in the air, natural camera motion",
"video_url": "https://example.com/input.mp3"
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());The request returns a generation id. Poll GET https://api.aimlapi.com/v2/video/generations?generation_id=<id> until the status is completed.
Models similar to Wan 2.2 Vace Inpainting
Related articles
Wan 2.2 Vace Inpainting FAQ
What inputs and outputs does Wan 2.2 Vace Inpainting support?
Wan 2.2 Vace Inpainting takes text, video as input and returns video.
How much does Wan 2.2 Vace Inpainting cost?
Wan 2.2 Vace Inpainting costs $0.05 per generation on AI/ML API, billed as you go with no subscription.
How is Wan 2.2 Vace Inpainting billed?
Wan 2.2 Vace Inpainting is billed per generation, a fixed charge per output rather than by prompt length.
Who created Wan 2.2 Vace Inpainting?
Wan 2.2 Vace Inpainting was built by Alibaba Cloud.
How do I call Wan 2.2 Vace Inpainting on AI/ML API?
Send a request to v2/video/generations with wan-22-vace-fun-a14b/inpainting as the model id.
Is Wan 2.2 Vace Inpainting available through a single API key?
Yes. Wan 2.2 Vace Inpainting is served through AI/ML API, so the same key and endpoint format used for other models applies.
What category of model is this?
It is classified as a video generation model with video editing features.
What is the Wan 2.2 Vace Inpainting model ID?
Use wan-22-vace-fun-a14b/inpainting as the model parameter. Requests go to https://api.aimlapi.com/v2/video/generations with your AI/ML API key.