Wan 2.1 I2v 720P

wan-2.1-i2v-720p

Accelerated inference for Wan 2.1 I2v 720P image to video with high resolution, a comprehensive and open suite of video foundation models that pushes the boundaries of video generation.

Fast Inference

REST API

Try in Console API Docs Examples

Model Information

Response Time~130 sec

StatusActive

Version

0.0.1

Updated3 days ago

API Reference

View Full Documentation

Prerequisites

Create an API Key from the Eachlabs Console
Install the required dependencies for your chosen language (e.g., requests for Python)

API Integration Steps

1. Create a Prediction

Send a POST request to create a new prediction. This will return a prediction ID that you'll use to check the result. The request should include your model inputs and API key.

import requests
import time

API_KEY = "YOUR_API_KEY"  # Replace with your API key
HEADERS = {
    "X-API-Key": API_KEY,
    "Content-Type": "application/json"
}

def create_prediction():
    response = requests.post(
        "https://api.eachlabs.ai/v1/prediction/",
        headers=HEADERS,
        json={
            "model": "wan-2-1-i2v-720p",
            "version": "0.0.1",
            "input": {
  "seed": 0,
  "image": "your image here",
  "prompt": "your prompt here",
  "max_area": "1280x720",
  "fast_mode": "Off",
  "num_frames": 81,
  "sample_shift": 5,
  "sample_steps": 30,
  "frames_per_second": 16,
  "sample_guide_scale": 5
}
        }
    )
    prediction = response.json()
    
    if prediction["status"] != "success":
        raise Exception(f"Prediction failed: {prediction}")
    
    return prediction["predictionID"]

2. Get Prediction Result

Poll the prediction endpoint with the prediction ID until the result is ready. The API uses long-polling, so you'll need to repeatedly check until you receive a success status.

def get_prediction(prediction_id):
    while True:
        result = requests.get(
            f"https://api.eachlabs.ai/v1/prediction/{prediction_id}",
            headers=HEADERS
        ).json()
        
        if result["status"] == "success":
            return result
        elif result["status"] == "error":
            raise Exception(f"Prediction failed: {result}")
        
        time.sleep(1)  # Wait before polling again

3. Complete Example

Here's a complete example that puts it all together, including error handling and result processing. This shows how to create a prediction and wait for the result in a production environment.

try:
    # Create prediction
    prediction_id = create_prediction()
    print(f"Prediction created: {prediction_id}")
    
    # Get result
    result = get_prediction(prediction_id)
    print(f"Output URL: {result['output']}")
    print(f"Processing time: {result['metrics']['predict_time']}s")
except Exception as e:
    print(f"Error: {e}")

Additional Information

The API uses a two-step process: create prediction and poll for results
Response time: ~130 seconds
Rate limit: 60 requests/minute
Concurrent requests: 10 maximum
Use long-polling to check prediction status until completion

Overview

Wan 2.1 I2V 720P is a model designed for generating high-quality videos from images based on textual descriptions. It supports frame-by-frame video generation with various customization options, enabling users to control the number of frames, resolution, sampling methods, and other parameters.

Technical Specifications

Optimization: Fine-tuned for generating smooth, natural-looking animations from static images

Use Case Suitability: Well-suited for animation prototyping, AI-assisted motion generation, and concept visualization

Processing Modes: Multiple settings (Off, Balanced, Fast, Ultra-fast) to optimize speed and quality

Training Data: Trained on high-quality image and motion datasets to ensure realistic frame transitions

Key Considerations

Wan 2.1 I2V 720P generates longer videos may require higher computation time and may impact consistency between frames.
Lower sample_steps values can speed up processing but may reduce detail in frames.
sample_guide_scale and sample_shift can significantly affect output quality; lower values maintain fidelity, while higher values introduce variations.
fast_mode settings affect processing time and quality trade-offs; use higher speeds only when necessary.

Tips & Tricks

Optimal Frame Settings: Use num_frames = 81 and frames_per_second = 16 for a good balance between length and smoothness.
Best Resolution Choice: Stick to 1280x720 or 720x1280 to avoid stretching or cropping artifacts.
Fine-tuning Sampling: Set sample_steps between 30-40 for detailed output; lower values speed up generation but reduce detail.
Adjusting Guidance Scale: For subtle refinements, use sample_guide_scale in the range of 4-7. Higher values can lead to exaggerated changes.
Using Fast Mode: If prioritizing quality, keep fast_mode at Balanced or Off; for quick drafts, Ultra-fast can be used.
Controlling Variability: sample_shift values between 3-7 offer a balance between stability and diversity in frame transitions.

Capabilities

with Wan 2.1 I2V 720P, you can convert static images into fluid motion sequences.
Supports different resolutions and frame rate configurations.
Provides adjustable sampling and guide settings for better control over the output.
Wan 2.1 I2V 720P can generate a variety of motion styles depending on input parameters.

What can I use for?

Animation Prototyping: Creating short animated clips from static images.
Content Creation: Enhancing illustrations or AI-generated art with movement.
Concept Visualization: Generating quick motion previews for storytelling or presentations.
AI-Assisted Creativity: Exploring new ways to animate characters, objects, and scenes.

Things to be aware of

Experiment with sample_steps = 35 and sample_guide_scale = 5 for a refined balance of detail and efficiency.
Use different fast_mode settings to compare speed vs. quality trade-offs.
Modify seed values to generate different variations of the same prompt.
Try varying num_frames between 40-81 to test different video lengths.
Adjust sample_shift values to introduce subtle motion variations for more dynamic results.

Limitations

Wan 2.1 I2V 720P may struggle with extreme motion consistency in long sequences.
High sample_guide_scale values may lead to unnatural artifacts.
Output quality depends on the clarity of the input image; low-quality inputs may produce less desirable results.
Processing time increases with higher frame counts and detailed sampling settings.

Output Format: MP4

Related AI Models