Wan 2.2 Image-to-Video Lora API by Alibaba
Open and Advanced Large-Scale Video Generative Models.
INPUT
OUTPUT
IdleYour request will cost $0.04 per run. For $10 you can run this model approximately 250 times.
Here's what you can do next:
Code Example
import requests
import time
# Step 1: Start video generation
generate_url = "https://api.atlascloud.ai/api/v1/model/generateVideo"
headers = {
"Content-Type": "application/json",
"Authorization": "Bearer $ATLASCLOUD_API_KEY"
}
data = {
"model": "atlascloud/wan-2.2/image-to-video-lora",
"prompt": "A beautiful sunset over the ocean with gentle waves",
"width": 512,
"height": 512,
"duration": 3,
"fps": 24,
}
generate_response = requests.post(generate_url, headers=headers, json=data)
generate_result = generate_response.json()
prediction_id = generate_result["data"]["id"]
# Step 2: Poll for result
poll_url = f"https://api.atlascloud.ai/api/v1/model/prediction/{prediction_id}"
def check_status():
while True:
response = requests.get(poll_url, headers={"Authorization": "Bearer $ATLASCLOUD_API_KEY"})
result = response.json()
if result["data"]["status"] in ["completed", "succeeded"]:
print("Generated video:", result["data"]["outputs"][0])
return result["data"]["outputs"][0]
elif result["data"]["status"] == "failed":
raise Exception(result["data"]["error"] or "Generation failed")
else:
# Still processing, wait 2 seconds
time.sleep(2)
video_url = check_status()Install
Install the required package for your language.
pip install requestsAuthentication
All API requests require authentication via an API key. You can get your API key from the Atlas Cloud dashboard.
export ATLASCLOUD_API_KEY="your-api-key-here"HTTP Headers
import os
API_KEY = os.environ.get("ATLASCLOUD_API_KEY")
headers = {
"Content-Type": "application/json",
"Authorization": f"Bearer {API_KEY}"
}Never expose your API key in client-side code or public repositories. Use environment variables or a backend proxy instead.
Submit a request
import requests
url = "https://api.atlascloud.ai/api/v1/model/generateVideo"
headers = {
"Content-Type": "application/json",
"Authorization": "Bearer $ATLASCLOUD_API_KEY"
}
data = {
"model": "your-model",
"prompt": "A beautiful landscape"
}
response = requests.post(url, headers=headers, json=data)
print(response.json())Submit a Request
Submit an asynchronous generation request. The API returns a prediction ID that you can use to check the status and retrieve the result.
/api/v1/model/generateVideoRequest Body
import requests
url = "https://api.atlascloud.ai/api/v1/model/generateVideo"
headers = {
"Content-Type": "application/json",
"Authorization": "Bearer $ATLASCLOUD_API_KEY"
}
data = {
"model": "atlascloud/wan-2.2/image-to-video-lora",
"input": {
"prompt": "A beautiful sunset over the ocean with gentle waves"
}
}
response = requests.post(url, headers=headers, json=data)
result = response.json()
print(f"Prediction ID: {result['id']}")
print(f"Status: {result['status']}")Response
{
"id": "pred_abc123",
"status": "processing",
"model": "model-name",
"created_at": "2025-01-01T00:00:00Z"
}Check Status
Poll the prediction endpoint to check the current status of your request.
/api/v1/model/prediction/{prediction_id}Polling Example
import requests
import time
prediction_id = "pred_abc123"
url = f"https://api.atlascloud.ai/api/v1/model/prediction/{prediction_id}"
headers = { "Authorization": "Bearer $ATLASCLOUD_API_KEY" }
while True:
response = requests.get(url, headers=headers)
result = response.json()
status = result["data"]["status"]
print(f"Status: {status}")
if status in ["completed", "succeeded"]:
output_url = result["data"]["outputs"][0]
print(f"Output URL: {output_url}")
break
elif status == "failed":
print(f"Error: {result['data'].get('error', 'Unknown')}")
break
time.sleep(3)Status Values
processingThe request is still being processed.completedGeneration is complete. Outputs are available.succeededGeneration succeeded. Outputs are available.failedGeneration failed. Check the error field.Completed Response
{
"data": {
"id": "pred_abc123",
"status": "completed",
"outputs": [
"https://storage.atlascloud.ai/outputs/result.mp4"
],
"metrics": {
"predict_time": 45.2
},
"created_at": "2025-01-01T00:00:00Z",
"completed_at": "2025-01-01T00:00:10Z"
}
}Upload Files
Upload files to Atlas Cloud storage and get a URL you can use in your API requests. Use multipart/form-data to upload.
/api/v1/model/uploadMediaUpload Example
import requests
url = "https://api.atlascloud.ai/api/v1/model/uploadMedia"
headers = { "Authorization": "Bearer $ATLASCLOUD_API_KEY" }
with open("image.png", "rb") as f:
files = {"file": ("image.png", f, "image/png")}
response = requests.post(url, headers=headers, files=files)
result = response.json()
download_url = result["data"]["download_url"]
print(f"File URL: {download_url}")Response
{
"data": {
"download_url": "https://storage.atlascloud.ai/uploads/abc123/image.png",
"file_name": "image.png",
"content_type": "image/png",
"size": 1024000
}
}Input Schema
The following parameters are accepted in the request body.
No parameters available.
Example Request Body
{
"model": "atlascloud/wan-2.2/image-to-video-lora"
}Output Schema
The API returns a prediction response with the generated output URLs.
Example Response
{
"id": "pred_abc123",
"status": "completed",
"model": "model-name",
"outputs": [
"https://storage.atlascloud.ai/outputs/result.mp4"
],
"metrics": {
"predict_time": 45.2
},
"created_at": "2025-01-01T00:00:00Z",
"completed_at": "2025-01-01T00:00:10Z"
}Atlas Cloud Skills
Atlas Cloud Skills integrates 300+ AI models directly into your AI coding assistant. One command to install, then use natural language to generate images, videos, and chat with LLMs.
Supported Clients
Install
npx skills add AtlasCloudAI/atlas-cloud-skillsSetup API Key
Get your API key from the Atlas Cloud dashboard and set it as an environment variable.
export ATLASCLOUD_API_KEY="your-api-key-here"Capabilities
Once installed, you can use natural language in your AI assistant to access all Atlas Cloud models.
MCP Server
Atlas Cloud MCP Server connects your IDE with 300+ AI models via the Model Context Protocol. Works with any MCP-compatible client.
Supported Clients
Install
npx -y atlascloud-mcpConfiguration
Add the following configuration to your IDE's MCP settings file.
{
"mcpServers": {
"atlascloud": {
"command": "npx",
"args": [
"-y",
"atlascloud-mcp"
],
"env": {
"ATLASCLOUD_API_KEY": "your-api-key-here"
}
}
}
}Available Tools
API Schema
Schema not availableNo examples available
Please log in to view request history
You need to be logged in to access your model request history.
Log InWan 2.2: Open and Advanced Large-Scale Video Generative Model by Alibaba Wanxiang
Model Card Overview
| Field | Description |
|---|---|
| Model Name | Wan 2.2 Image-to-Video LoRA |
| Developed by | Alibaba Tongyi Wanxiang Lab |
| Model Type | Image-to-Video Generation with LoRA Support |
| Resolution | 480p, 720p (via VSR upscaling) |
| Frame Rate | 30 fps |
| Duration | 3–10 seconds |
| Related Links | GitHub: https://github.com/Wan-Video/Wan2.2, Hugging Face: https://huggingface.co/Wan-AI/Wan2.2-I2V-A14B, Paper (arXiv): https://arxiv.org/abs/2503.20314 |
Introduction
Wan 2.2 is a significant upgrade to the Wan series of foundational video models, designed to push the boundaries of generative AI in video creation. This image-to-video LoRA variant takes a reference image as the first frame and generates a high-quality video, with full support for custom LoRA weights to fine-tune the generation style, motion characteristics, or subject identity.
The model generates videos at 480p natively and supports 720p output via Video Super Resolution (VSR) upscaling, delivering smooth 30 fps playback at both resolutions.
Key Features & Innovations
-
Effective MoE Architecture: Wan 2.2 integrates a Mixture-of-Experts (MoE) architecture into the video diffusion model. Specialized expert models handle different stages of the denoising process, increasing model capacity without raising computational costs. The model has 27B total parameters with only 14B active during any given step.
-
Cinematic-Level Aesthetics: Trained on a meticulously curated dataset with detailed labels for cinematic properties like lighting, composition, and color tone. This allows generation of videos with precise and controllable artistic styles, achieving a professional, cinematic look.
-
Complex Motion Generation: Trained on a vastly expanded dataset (+65.6% more images and +83.2% more videos compared to Wan 2.1), Wan 2.2 demonstrates superior ability to generate complex and realistic motion with enhanced generalization across motions, semantics, and aesthetics.
-
Custom LoRA Support: This variant supports user-provided LoRA weights for fine-grained style and motion control. Three separate LoRA input channels are available:
high_noise_loras— Applied to the high-noise expert (transformer stage), influencing overall structure and layout.low_noise_loras— Applied to the low-noise expert (transformer_2 stage), influencing fine details and textures.loras— General-purpose LoRA input where the module is auto-inferred from the safetensors filename.
-
VSR-Enhanced Output: All output videos are delivered at 30 fps. When 720p resolution is selected, the model leverages Video Super Resolution to upscale from a 480p base generation, preserving fine details while achieving higher resolution output.
Model Architecture
The architecture is built upon the Diffusion Transformer (DiT) paradigm with a Mixture-of-Experts (MoE) framework:
- High-Noise Expert: Activated during initial denoising stages, establishing overall structure and layout.
- Low-Noise Expert: Activated in later stages, refining details, textures, and fine-grained motion.
The transition between experts is dynamically determined by the signal-to-noise ratio (SNR) during generation. Custom LoRA weights can be applied to each expert independently, enabling precise control over different aspects of the generation pipeline.
Intended Use & Applications
- Stylized Video Production: Generating videos with custom visual styles by applying LoRA weights trained on specific aesthetic data.
- Character & Subject Consistency: Using identity-preserving LoRAs to maintain consistent characters across multiple video generations.
- Cinematic Video Production: Generating high-fidelity video clips from reference images for short films, advertisements, or social media content.
- Creative Experimentation: Combining multiple LoRAs to explore novel visual effects and motion styles.
- Academic Research: Serving as a powerful foundation model for researchers exploring LoRA-based fine-tuning techniques in video generation.
