Generate Images from Text Prompts
MCP server that generates images from text prompts via Replicate's Flux model, with configurable aspect ratio, format, and output count.
Why it matters
Leverage the Replicate Flux model to generate high-quality images from text descriptions. Customize aspect ratios, output formats, and reproducibility for diverse creative needs.
Outcomes
What it gets done
Generate images based on text prompts.
Customize image aspect ratios and output formats.
Create multiple image variations with seed control.
Integrate image generation directly into Claude Desktop.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/vb-image-generation | bash Capabilities
Tools your agent gets
Generates images using the Flux model based on text prompts with customizable parameters like aspect ratio and seed.
Overview
Image Generation MCP Server
An MCP server with one tool, generate_image, that produces images from text prompts via Replicate's Flux model, returning URLs to the results. Use it when you want an AI assistant to generate images on demand through your own Replicate account, without building a custom integration.
What it does
Image Generation MCP Server exposes one tool, generate_image, that turns a text prompt into one or more images using Replicate's Flux model - defaulting to black-forest-labs/flux-schnell, overridable via a MODEL environment variable. It returns an array of URLs to the generated images rather than raw image bytes.
When to use - and when NOT to
Use it when you want an AI client to generate images from a prompt without writing custom Replicate API integration code. It requires your own Replicate account and a REPLICATE_API_TOKEN - there is no local or offline generation path, every call goes out to Replicate's hosted API and is billed on that account. The upstream repository (GongRzhe/Image-Generation-MCP-Server on GitHub) is archived and no longer receiving updates.
Capabilities
generate_image takes a required prompt (text description of the desired image) plus optional parameters: seed (for reproducible generation), aspect_ratio (default 1:1), output_format (webp, jpg, or png, default webp), and num_outputs (1-4 images, default 1). The autoApprove config array controls which tool names can run without a user confirmation prompt - empty by default, meaning every generate_image call requires confirmation.
How to install
Install automatically for Claude Desktop via Smithery:
npx -y @smithery/cli install @GongRzhe/Image-Generation-MCP-Server --client claude
Or run it directly with npx (no local install needed), or install it globally/locally with npm install -g @gongrzhe/image-gen-server (or without -g for a local copy). For Claude Desktop, add it to the mcpServers section of the config file (~/Library/Application Support/Claude/claude_desktop_config.json on macOS, %APPDATA%/Claude/claude_desktop_config.json on Windows) - either the recommended npx form (command: npx, args: ["@gongrzhe/image-gen-server"]) or, for a local install, command: node with args pointing at the built build/index.js path. Both forms need REPLICATE_API_TOKEN and optionally MODEL set under env, plus a disabled flag. Get a Replicate token by signing up or logging in at replicate.com, opening replicate.com/account/api-tokens, creating a new token, then pasting it in place of the placeholder value in the MCP config. The project is licensed under the MIT License.
Who it's for
Developers who want their AI assistant to generate images on request using an existing Replicate account, without building their own Replicate client or hosting a generation pipeline themselves. The disabled flag in the config gives a quick way to turn the server off without removing it entirely.
const result = await use_mcp_tool({
server_name: "image-gen",
tool_name: "generate_image",
arguments: {
prompt: "A beautiful sunset over mountains",
aspect_ratio: "16:9",
output_format: "png",
num_outputs: 1
}
});
Source README
Image Generation MCP Server
This MCP server provides image generation capabilities using the Replicate Flux model.
Installation
Installing via Smithery
To install Image Generation MCP Server for Claude Desktop automatically via Smithery:
npx -y @smithery/cli install @GongRzhe/Image-Generation-MCP-Server --client claude
Option 1: NPX Method (No Local Setup Required)
You can use the package directly from npm without installing it locally:
# No installation needed - npx will handle it
Option 2: Local Installation
If you prefer a local installation:
# Global installation
npm install -g @gongrzhe/image-gen-server
# Or local installation
npm install @gongrzhe/image-gen-server
Setup
Configure Claude Desktop
Edit your Claude Desktop configuration file:
- On MacOS:
~/Library/Application Support/Claude/claude_desktop_config.json - On Windows:
%APPDATA%/Claude/claude_desktop_config.json
Option 1: NPX Configuration (Recommended)
This method runs the server directly from npm without needing local files:
{
"mcpServers": {
"image-gen": {
"command": "npx",
"args": ["@gongrzhe/image-gen-server"],
"env": {
"REPLICATE_API_TOKEN": "your-replicate-api-token",
"MODEL": "alternative-model-name"
},
"disabled": false,
"autoApprove": []
}
}
}
Option 2: Local Installation Configuration
If you installed the package locally:
{
"mcpServers": {
"image-gen": {
"command": "node",
"args": ["/path/to/image-gen-server/build/index.js"],
"env": {
"REPLICATE_API_TOKEN": "your-replicate-api-token",
"MODEL": "alternative-model-name"
},
"disabled": false,
"autoApprove": []
}
}
}
Get Your Replicate API Token
- Sign up/login at https://replicate.com
- Go to https://replicate.com/account/api-tokens
- Create a new API token
- Copy the token and replace
your-replicate-api-tokenin the MCP settings
Environment Variables
REPLICATE_API_TOKEN(required): Your Replicate API token for authenticationMODEL(optional): The Replicate model to use for image generation. Defaults to "black-forest-labs/flux-schnell"
Configuration Parameters
disabled: Controls whether the server is enabled (false) or disabled (true)autoApprove: Array of tool names that can be executed without user confirmation. Empty array means all tool calls require confirmation.
Available Tools
generate_image
Generates images using the Flux model based on text prompts.
Parameters
prompt(required): Text description of the image to generateseed(optional): Random seed for reproducible generationaspect_ratio(optional): Image aspect ratio (default: "1:1")output_format(optional): Output format - "webp", "jpg", or "png" (default: "webp")num_outputs(optional): Number of images to generate (1-4, default: 1)
Example Usage
const result = await use_mcp_tool({
server_name: "image-gen",
tool_name: "generate_image",
arguments: {
prompt: "A beautiful sunset over mountains",
aspect_ratio: "16:9",
output_format: "png",
num_outputs: 1
}
});
The tool returns an array of URLs to the generated images.
๐ License
This project is licensed under the MIT License.
FAQ
Common questions
Discussion
Questions & comments ยท 0
Sign In Sign in to leave a comment.