AI Image Editor

AI-powered image editing using text prompts. Transform photos, product images, and artwork with natural language instructions.

What it does

This actor edits images based on a text prompt using the RunPod qwen-image-edit-2511-lora endpoint. Upload an image, describe the changes you want, and receive an edited version.

• 🖼️ Edit any image via URL using AI-powered image-to-image transformation
• 📝 Natural language prompts — describe what you want changed
• ⚡ Synchronous execution — waits for the result and returns it
• 🎨 Supports JPEG, PNG, and WebP output formats
• 🔧 Optional LoRA adapters and advanced parameters

Input

{
    "imageUrl": "https://example.com/photo.jpg",
    "prompt": "A futuristic city with glowing neon lights",
    "outputFormat": "jpeg",
    "aspectRatio": "1024*1024",
    "seed": -1
}

Input fields

Field	Type	Required	Description
imageUrl	string	✅	URL of the image to edit
prompt	string	✅	Text description of the desired edit
outputFormat	string	❌	Output format: jpeg, png, webp (default: jpeg)
aspectRatio	string	❌	Output dimensions: 1024*1024, 1024*1280, 1280*1024, 1280*1280, 1280*1536, 1536*1080 (default: 1024*1024)
seed	integer	❌	Random seed for reproducibility. -1 = random (default: -1)
numInferenceSteps	integer	❌	Number of denoising steps (1–100)
guidance	number	❌	Guidance scale for prompt adherence (1–20)
enableBase64Output	boolean	❌	Return image as base64 in response (default: false)

Output

The actor produces two outputs:

📊 Dataset — Run Results

Metadata about the editing job, including a public URL to the edited image:

{
    "status": "COMPLETED",
    "prompt": "A futuristic city with glowing neon lights",
    "originalImageUrl": "https://example.com/photo.jpg",
    "outputFormat": "jpeg",
    "imageSize": 245760,
    "imageExtension": "jpg",
    "contentType": "image/jpeg",
    "publicImageUrl": "https://api.apify.com/v2/key-value-stores/xxx/records/edited-image",
    "delayTime": 1500,
    "executionTime": 8200,
    "runpodJobId": "abc-123"
}

🖼️ Key-Value Store — Edited Image

The generated/edited image is saved as edited-image in the default key-value store. It is accessible via a public URL that is returned in the dataset results.

Use Cases

• 👗 Fashion & e-commerce — change backgrounds, lighting, or styles on product photos
• 🏠 Real estate — enhance property images, change time of day, add furnishings
• 🎨 Creative workflows — iterate on concept art and illustrations
• 📸 Photography — apply stylistic edits, lighting corrections, or scene modifications
• 🛍️ Marketing — quickly generate variations of campaign imagery

Tips & Best Practices

• Be specific in prompts — detailed descriptions yield better edits
• Use seeds — set a fixed seed to reproduce the same edit style across multiple images
• Start with defaults — the default 1024*1024 size and format work well for most cases
• Check RunPod dashboard — if jobs fail, verify your API key and endpoint availability
• Timeout — the actor waits up to 5 minutes for RunPod to complete the job