Stable Diffusion Inpainting

stable-diffusion-inpainting

Stable Diffusion Inpainting is a model that can be used to generate and modify images based on text prompts.

A100 80GB

Fast Inference

REST API

Try in Console API Docs Examples

Model Information

Response Time~1 sec

StatusActive

Version

0.0.1

Updated20 days ago

Live Demo

Average runtime: ~1 seconds

Input

Configure model parameters

Negative Prompt

Specify things to not see in the output

Enter your negative prompt

Prompt

Input prompt

Face of a yellow rabbit, high resolution, sitting on a park bench

Image

Initial image to generate variations of. Will be resized to height x width

File upload is currently disabled

JPEGPNGJPGWEBP

Mask

Black and white image to use as mask for inpainting over the image provided. White pixels are inpainted and black pixels are preserved.

File upload is currently disabled

JPEGPNGJPGWEBP

height

An enumeration.

Enter your height

width

An enumeration.

Enter your width

Guidance Scale

Scale for classifier-free guidance

7.5

scheduler

An enumeration.

Seed

Random seed. Leave blank to randomize the seed

Enter your seed

Output

View generated results

Result

Preview, share or download your results with a single click.

Cost is calculated based on execution time.The model is charged at $0.002 per second. With a $1 budget, you can run this model approximately 500 times, assuming an average execution time of 1 seconds per run.

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": "stable-diffusion-inpainting",
            "version": "0.0.1",
            "input": {
  "mask": "your_file.image/jpeg",
  "seed": null,
  "image": "your_file.image/jpeg",
  "width": "512",
  "height": "512",
  "prompt": "a vision of paradise. unreal engine",
  "scheduler": "DPMSolverMultistep",
  "num_outputs": "1",
  "guidance_scale": "7.5",
  "negative_prompt": "your negative prompt here",
  "num_inference_steps": "50",
  "disable_safety_checker": false
}
        }
    )
    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: ~1 seconds
Rate limit: 60 requests/minute
Concurrent requests: 10 maximum
Use long-polling to check prediction status until completion

Overview

Stable Diffusion Inpainting is designed for inpainting tasks, enabling users to seamlessly restore or modify images by filling in missing or altered parts. Leveraging textual descriptions and visual cues, the model produces outputs that integrate naturally with the existing image. It is ideal for tasks like restoration, creative editing, and customized visual projects, offering numerous parameters for precise control.

Technical Specifications

Inpainting Capability: The model specializes in filling masked areas of an image with content that aligns with the provided prompt while maintaining consistency with the surrounding context.

High-Quality Output: Generates detailed and coherent images with realistic textures, lighting, and blending.

Versatile Modifications: Supports a wide range of image edits, including object addition, background changes, and defect correction.

Prompt-Driven Generation: Relies on descriptive text prompts to guide the inpainting process, allowing for creative and specific outputs.

Consistent Reproducibility: Allows for deterministic results through seed-based random number generation.

Content Safety Features: Includes an optional safety checker to ensure outputs adhere to appropriate content guidelines.

Key Considerations

Image Quality: Use high-resolution base images for optimal results.
Mask Precision: Ensure the mask accurately defines the area to modify.
Prompt-Output Alignment: Avoid overly complex or conflicting prompts that may confuse the model.
Inference Steps: Balance between time and detail; excessive steps may not always yield noticeable improvements.

Tips & Tricks

Refining Prompts: Start with a general description and gradually add details for better control.
Negative Prompting: Use the negative_prompt parameter to exclude unwanted elements effectively.
Custom Resolutions: Match aspect_ratio to the dimensions of your input image for consistency.

Capabilities

Realistic Inpainting: Seamlessly integrates new elements into existing images.
Customizable Outputs: Adjust parameters to meet specific project needs.

What can I use for?

Image Restoration: Repair damaged or incomplete photos and artwork.
Creative Edits: Add, modify, or remove elements for artistic purposes.
Design Enhancements: Tailor visuals for presentations, media, or marketing.
Prototype Visualization: Generate concept visuals quickly for brainstorming.

Things to be aware of

Restoration Projects: Repair old or damaged images by filling in missing parts.
Creative Variations: Explore different prompts with the same base image to generate unique outputs.
Custom Masking: Define intricate areas to edit for precise results.
Size Testing: Compare results at different sizes to determine the best settings for your use case.

SCHEDULER

DDIM:
- Use for faster results with smooth transitions.
- Works well with fewer inference steps.
K_EULER:
- Ideal for sharp and detailed outputs.
- Pair with medium to high guidance scale values for clarity.
DPMSolverMultistep:
- Best for balancing speed and quality.
- Delivers excellent results with fewer steps.
K_EULER_ANCESTRAL:
- Great for creative and artistic outputs.
- Experiment with lower guidance scales for diverse results.
PNDM:
- Ensures high accuracy and consistency.
- Use more steps for highly detailed outputs.
KLMS:
- Perfect for high-resolution, realistic images.
- Higher guidance scales enhance detailed scenes like landscapes.

Limitations

Output Consistency: Complex prompts or conflicting inputs may lead to unpredictable results.
Resolution Constraints: Extremely high resolutions may increase processing time significantly.
Mask Limitations: Poorly defined masks can lead to unintended modifications.
Safety Checker: Disabling it may result in outputs that do not meet content guidelines.

Output Format: PNG

Related AI Models

Illusion Diffusion

illusion-diffusion-hq

Image to Image

Style Changer by ByteDance

bytedance

Image to Image

Flux Redux Dev

flux-redux-dev

Image to Image

Depth Anything

depth-anything

Image to Image