Skip to content
Back to skills

Image Merger

ASecurity

Image Merger | AI Image Blending. Merge and blend two images together with multiple merge modes and optional animation. Triggers: image merger, merge images, blend images, image blending, combine images

  • 29 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added June 2, 2026
documentationbashdebuggingapidocumentation

Works with

  • api

Security analysis

A100/100

Scanned June 2, 2026

npx -y skills add awesome-genmedia/skills --skill image-merger --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Image Merger?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Image Merger
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/awesome-genmedia-image-merger/badge)](https://www.skillsdirectory.com/skills/awesome-genmedia-image-merger)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: image-merger
description: "Image Merger | AI Image Blending. Merge and blend two images together with multiple merge modes and optional animation. Triggers: image merger, merge images, blend images, image blending, combine images"
allowed-tools: Bash(curl *), WebFetch
---

# Image Merger

Merge and blend two images together with multiple merge modes, optional ControlNet guidance, upscaling, and animation support. Create seamless composites from separate source images.

## Quick Start

> Requires an each::labs API key. Get one at [eachlabs.ai](https://eachlabs.ai).

```bash
curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "image-merger",
    "version": "0.0.1",
    "input": {
      "image_1": "https://example.com/landscape.jpg",
      "image_2": "https://example.com/portrait.jpg",
      "prompt": "a person standing in a beautiful mountain landscape",
      "merge_mode": "full",
      "steps": 20
    }
  }'
```

## Parameters

| Parameter | Type | Default | Description |
|-----------|------|---------|-------------|
| animate | boolean | false | Animate merging from one image to the other (returns video) |
| animate_frames | integer | 24 | Number of frames for the animation |
| control_image | string | - | Optional image to influence merging with ControlNet |
| height | integer | 768 | Output height |
| image_1 | string | - | First image |
| image_1_strength | number | 1 | Strength of the first image |
| image_2 | string | - | Second image |
| image_2_strength | number | 1 | Strength of the second image |
| merge_mode | string | full | Merge mode. enum: full, left_right, top_bottom |
| negative_prompt | string | ugly, broken, distorted | Things to avoid in the merged image |
| prompt | string | a photo | Prompt to guide image merging |
| return_temp_files | boolean | false | Return preprocessed controlnet images for debugging |
| seed | integer | - | Fix random seed for reproducibility |
| steps | integer | 20 | Number of steps |
| upscale_2x | boolean | false | Upscale output 2x |
| upscale_steps | integer | 20 | Steps per controlnet tile for upscaling |
| width | integer | 768 | Output width |

## Examples

**Split merge (left-right):**
```bash
curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "image-merger",
    "version": "0.0.1",
    "input": {
      "image_1": "https://example.com/summer-scene.jpg",
      "image_2": "https://example.com/winter-scene.jpg",
      "prompt": "a landscape transitioning from summer to winter, seamless blend",
      "merge_mode": "left_right",
      "image_1_strength": 1,
      "image_2_strength": 1,
      "steps": 25
    }
  }'
```

**Animated merge transition:**
```bash
curl -X POST https://api.eachlabs.ai/v1/prediction \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "model": "image-merger",
    "version": "0.0.1",
    "input": {
      "image_1": "https://example.com/young-portrait.jpg",
      "image_2": "https://example.com/old-portrait.jpg",
      "prompt": "a portrait of a person aging gracefully",
      "animate": true,
      "animate_frames": 30,
      "steps": 20,
      "seed": 42
    }
  }'
```

## Related Models

- [become-image](../become-image/) - Transform a person into any style
- [face-swap-new](../face-swap-new/) - Face swapping between images
- [flux-fill-pro](../flux-fill-pro/) - Inpainting and image editing

## Documentation

- [each::labs Docs](https://docs.eachlabs.ai)
- [API Reference](https://docs.eachlabs.ai/api/overview)

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.