generate_ai_imageLABSCO SUMMARY
Text-to-image generation inline, with size presets for the common aspect ratios.
The shortcut tools remove the one parameter people fumble most — the exact pixel size — so an agent can ask for a landscape image and get a sensible aspect ratio, while generate_ai_image stays available when the size or model needs to be specified exactly.
What it is
A server that wraps an image-generation library over OpenAI and Replicate, so an assistant can produce images from a text prompt with control over size, model, and output path.
What you get
- generate_ai_image creates an image from a prompt, with optional size, model, and output-file path
- square_image, landscape_image, and portrait_image are preset-size shortcuts that take the same parameters without needing a size
- Built on the ai-image module for OpenAI and Replicate image models
Requirements
API keys for OpenAI and/or Replicate, supplied in the client configuration; runs through npx on Node.
Setup effort
One command plus a key — npm install -g image-mcp, then supply credentials
USE CASES
01
Making an image from a description
generate_ai_image turns a text prompt into a picture, with model and size configurable when it matters.
02
Getting the aspect ratio right without thinking about pixels
landscape_image or portrait_image applies a preset size so the result matches the intended orientation.
03
Producing a square asset
square_image generates a square image for an avatar or icon slot.
4 tools
Listed in project docs. Your client may show a different list.
square_imagelandscape_imageportrait_imageBEFORE YOU INSTALL
What to know first
Tools you get
4 tools named in the project's documentation.
Requires
- npx on PATH
- Node.js >= 18.0.0
- OPENAI_API_KEY or REPLICATE_API_TOKEN (one of the two, for the provider whose models you use)
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
TERMINAL
npm install -g image-mcpMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"image-mcp": {
"command": "npx",
"args": [
"image-mcp@latest"
],
"env": {
"OPENAI_API_KEY": "<YOUR_OPENAI_API_KEY>",
"REPLICATE_API_TOKEN": "<YOUR_REPLICATE_API_TOKEN>",
"MCP_TIMEOUT": "1200000"
},
"resetTimeoutOnProgress": true
}
}
}Keep the key for the provider whose models you use and delete the other line; the project asks for an OpenAI and/or a Replicate key, not both.
