Generates post images in Facebook-optimized dimensions (1200x630).
Generates featured images and banners in Ghost-optimized dimensions (1200x675, 2000x1125, 2560x1440) for blog posts.
Generates blog and social media images using Google's Gemini AI image generation models with configurable quality levels and platform-specific presets.
Generates images optimized for Instagram posts (1080x1080 square) and stories/reels (1080x1920 vertical).
Generates banner images in Medium-optimized dimensions (1400x788) for articles.
Generates header images in Substack-optimized dimensions (1456x816) for newsletter posts.
Generates featured images in WordPress-optimized dimensions (1200x675) for blog posts.
Generates thumbnails (1280x720) and channel banners (2560x1440) optimized for YouTube.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Image Generation MCP create a LinkedIn post image about AI ethics with a modern minimalist style"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
An MCP (Model Context Protocol) server for generating blog and social media images using AI. Currently supports Google's Gemini/Nano Banana image generation models with a provider architecture designed for easy extension.
Install: npx -y image-generation-mcp
Platform Presets: Pre-configured dimensions for Ghost, Medium, Instagram, Twitter, LinkedIn, YouTube, and more
Multiple Quality Levels: Standard (fast) or High (uses Gemini Pro for better quality)
Auto-Save: Images are always saved to disk (never lost as base64-only responses)
PNG Metadata: Prompt, model, style, and generation info embedded in every image
Provider Architecture: Extensible design to support multiple AI providers
Security First: Input validation, prompt sanitization, safe error handling
Related MCP server: Gemini Image MCP Server
Variable | Required | Description |
| Yes | Your Google AI API key for Gemini |
| No | Default directory for saved images (defaults to |
Get your API key from Google AI Studio.
# Add to Claude Code (user scope - available in all projects)
claude mcp add image-gen --scope user -e GOOGLE_API_KEY=your-api-key -- npx -y image-generation-mcp
# Or add to current project only
claude mcp add image-gen -e GOOGLE_API_KEY=your-api-key -- npx -y image-generation-mcp# From the project directory, build first
npm run build
# Add local server to Claude Code
claude mcp add image-gen -e GOOGLE_API_KEY=your-api-key -- node /absolute/path/to/image-generation-mcp/dist/index.jsAdd to ~/.claude.json (user scope) or .mcp.json (project scope):
{
"mcpServers": {
"image-gen": {
"type": "stdio",
"command": "npx",
"args": ["-y", "image-generation-mcp"],
"env": {
"GOOGLE_API_KEY": "your-api-key-here"
}
}
}
}For local development:
{
"mcpServers": {
"image-gen": {
"type": "stdio",
"command": "node",
"args": ["/absolute/path/to/image-generation-mcp/dist/index.js"],
"env": {
"GOOGLE_API_KEY": "your-api-key-here"
}
}
}
}# List configured MCP servers
claude mcp list
# Check status within Claude Code
/mcpAdd to your claude_desktop_config.json:
{
"mcpServers": {
"image-gen": {
"command": "npx",
"args": ["-y", "image-generation-mcp"],
"env": {
"GOOGLE_API_KEY": "your-api-key-here"
}
}
}
}Config file locations:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
Generate an image for blog posts or social media.
Parameters:
Parameter | Type | Required | Description |
| string | Yes | Description of the image to generate |
| string | No | Platform preset (default: |
| string | No |
|
| string | No | Style hint (e.g., "photorealistic", "illustration") |
| string | No | Blog post title for context |
| string | No | Path to save the image (defaults to |
| string | No | Provider to use (default: |
Example:
Generate a blog banner for my post about TypeScript best practices.
Use format: medium-ghost-spooky, style: modern minimalistList all available image format presets.
Parameters:
Parameter | Type | Required | Description |
| string | No | Filter by category: |
Format | Dimensions | Ratio | Description |
| 1200x675 | 16:9 | Featured image for Ghost blog posts |
| 2000x1125 | 16:9 | High-resolution feature image for Ghost |
| 2560x1440 | 16:9 | Premium high-resolution blog banner (QHD) |
| 1400x788 | 16:9 | Banner image for Medium articles |
| 1456x816 | 16:9 | Header image for Substack posts |
| 1200x675 | 16:9 | Featured image for WordPress posts |
Format | Dimensions | Ratio | Description |
| 1080x1080 | 1:1 | Square post for Instagram feed |
| 1080x1920 | 9:16 | Vertical story/reel for Instagram |
| 1200x675 | 16:9 | Image for Twitter/X posts |
| 1200x628 | ~1.91:1 | Image for LinkedIn posts |
| 1200x630 | ~1.91:1 | Image for Facebook posts |
Format | Dimensions | Ratio | Description |
| 1280x720 | 16:9 | Thumbnail for YouTube videos |
| 2560x1440 | 16:9 | Channel banner for YouTube |
Format | Dimensions | Ratio | Description |
| 1024x1024 | 1:1 | Generic square image |
| 1920x1080 | 16:9 | Standard landscape (1080p) |
| 3840x2160 | 16:9 | 4K landscape image |
| 1080x1920 | 9:16 | Standard portrait/vertical image |
Every generated PNG image includes embedded metadata:
Field | Description |
| The original prompt |
| Model used (e.g., |
| Provider name ( |
| Preset used (e.g., |
| Style hint if specified |
| Blog post title if specified |
| ISO timestamp |
|
|
View metadata:
# macOS/Linux
strings your-image.png | grep -E "^(Description|AI-|Title|Creation)"
# Or use exiftool
exiftool your-image.pngThis MCP server implements several security measures:
Input Validation: Prompts are validated for length and sanitized
Prompt Injection Protection: Suspicious patterns are blocked
Path Traversal Prevention: Output paths are validated
Safe Error Messages: API keys and sensitive data are never exposed in errors
No Logging of Secrets: API keys are never logged
This is a simple, vibe-coded MCP server for generating images. It is provided as-is for convenience and educational purposes.
API Key Security: Your GOOGLE_API_KEY is as safe as you make it. We do not store, log, or transmit your API key anywhere except to Google's API. You are responsible for:
Keeping your API key secure
Not committing it to version control
Rotating it if you suspect it has been compromised
Data Transmission: Your prompts and generated images are sent to/from Google's Gemini API. Review Google's AI Terms of Service for their data handling policies.
No Warranty: This software is provided "AS IS", without warranty of any kind. The authors are not liable for any damages, data loss, API costs, or other issues arising from use of this software.
API Costs: Image generation may incur costs on your Google Cloud account. Monitor your usage and set up billing alerts.
Content Responsibility: You are responsible for the prompts you submit and the images you generate. Do not use this tool to generate harmful, illegal, or policy-violating content.
MIT License - see LICENSE for full terms.
By using this software, you acknowledge that you have read and understood these terms.
The server uses a provider interface pattern. To add a new provider:
Create a new file in src/providers/ implementing ImageProvider
Register it in src/providers/index.ts
// src/providers/my-provider.ts
import { ImageProvider, ImageGenerationOptions, GeneratedImage } from "./types.js";
export class MyProvider implements ImageProvider {
readonly name = "my-provider";
isConfigured(): boolean { /* ... */ }
generateImage(options: ImageGenerationOptions): Promise<GeneratedImage> { /* ... */ }
getSupportedAspectRatios(): string[] { /* ... */ }
getMaxResolution(): { width: number; height: number } { /* ... */ }
}# Install dependencies
npm install
# Build
npm run build
# Run locally
GOOGLE_API_KEY=your-key node dist/index.js
# Watch mode
npm run dev
# Run tests
npm test
# Lint and format
npm run lint
npm run format# 1. Make sure you're logged in to npm
npm login
# 2. Update version in package.json (if needed)
npm version patch # or minor, major
# 3. Run all checks
npm run check
# 4. Publish
npm publish
# 5. After publishing, users can install with:
# npx image-generation-mcp
# or: npm install -g image-generation-mcpMIT
Generate images for blog posts and social media using AI.
Available formats:
ghost-banner: Ghost Blog Banner (1200x675)
ghost-feature: Ghost Feature Image (HD) (2000x1125)
medium-ghost-spooky: Medium Ghost Spooky (2560x1440)
medium-banner: Medium Banner (1400x788)
substack-header: Substack Header (1456x816)
wordpress-featured: WordPress Featured (1200x675)
instagram-post: Instagram Post (1080x1080)
instagram-story: Instagram Story (1080x1920)
instagram-landscape: Instagram Landscape (1080x608)
twitter-post: Twitter/X Post (1200x675)
twitter-header: Twitter/X Header (1500x500)
linkedin-post: LinkedIn Post (1200x628)
linkedin-banner: LinkedIn Banner (1584x396)
facebook-post: Facebook Post (1200x630)
facebook-cover: Facebook Cover (820x312)
youtube-thumbnail: YouTube Thumbnail (1280x720)
youtube-banner: YouTube Banner (2560x1440)
square: Square (1024x1024)
square-hd: Square HD (2048x2048)
landscape: Landscape (1920x1080)
landscape-4k: Landscape 4K (3840x2160)
portrait: Portrait (1080x1920)
Examples:
Generate a Ghost blog banner: { "prompt": "A serene mountain landscape at sunset", "format": "ghost-banner" }
High quality Instagram post: { "prompt": "Minimalist coffee cup on marble", "format": "instagram-post", "quality": "high" }
YouTube thumbnail with title: { "prompt": "Exciting tech reveal", "format": "youtube-thumbnail", "title": "New iPhone 17 Review" }
IMPORTANT: Always specify an outputPath to save the image to a meaningful location. If omitted, images are saved to a generated-images/ directory in the current working directory with a timestamped filename.
| Name | Required | Description | Default |
|---|---|---|---|
| prompt | Yes | Description of the image to generate | |
| format | No | Platform preset: ghost-banner, ghost-feature, medium-ghost-spooky, medium-banner, substack-header, wordpress-featured, instagram-post, instagram-story, instagram-landscape, twitter-post, twitter-header, linkedin-post, linkedin-banner, facebook-post, facebook-cover, youtube-thumbnail, youtube-banner, square, square-hd, landscape, landscape-4k, portrait | ghost-banner |
| quality | No | Quality level | standard |
| style | No | Optional style hint (e.g., 'photorealistic', 'illustration') | |
| title | No | Optional blog post title for context | |
| outputPath | No | Optional path to save the image file | |
| provider | No | Image generation provider | gemini |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, description discloses default save behavior and outputPath importance. Does not mention auth, rate limits, or side effects beyond file creation, but cover key behavioral aspects adequately.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Well-organized with headings, bullet list, examples, and a note. Every section serves a purpose, though the format list could be condensed. Front-loaded with purpose, then details.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a 7-param tool with no output schema or annotations, the description explains all inputs, default behaviors, and provides examples. Lacks return value details but that falls on schema.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, baseline 3. Description adds dimensions to each format, examples for quality/title, and clarifies outputPath purpose. Adds meaningful context beyond schema definitions.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description clearly states 'Generate images for blog posts and social media using AI' - a specific verb and resource. Sibling tool 'list_image_formats' lists formats, while this generates images, making the distinction obvious.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Examples show typical use cases and the IMPORTANT note guides outputPath usage. Lacks explicit when-not-to-use or alternative tools, but the sibling count is small and context is clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List all available image format presets with their dimensions and use cases.
| Name | Required | Description | Default |
|---|---|---|---|
| category | No | Filter by category |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries full burden. It clearly indicates a read-only listing operation and describes the content of the output (dimensions and use cases). It does not discuss permissions or side effects, but for a simple listing tool, this is sufficient.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, concise sentence that front-loads the action ('List all...') and specifies what the output contains. Every word adds value, and there is no redundancy or filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the low complexity (1 optional parameter, no output schema), the description adequately covers the tool's purpose and output content. It could explicitly state that all presets are returned when no category is provided, but the word 'all' implies that. Sibling context is clear.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The input schema has 100% coverage with descriptions for the 'category' parameter, including an enum. The description does not add additional meaning to the parameter beyond what the schema provides, so baseline score of 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool lists all available image format presets and what they include (dimensions and use cases). It is distinct from the sibling 'generate_blog_image' which likely generates an image. However, it does not explicitly differentiate from siblings, missing a chance for clarity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies the tool is used to view available presets before generating an image, but provides no explicit guidance on when to use or not use it, nor any alternatives. The context of a single sibling suggests usage, but lacks direct instruction.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
generate_blog_imagelist_image_formatsScored across 2 tools
The two tools have clearly distinct purposes: one generates images with various format options, the other lists available formats. No overlap or ambiguity between them.
Both tools follow a consistent verb_noun pattern (generate_blog_image, list_image_formats), making it easy to understand their function at a glance.
With only two tools, the server feels slightly thin for its domain. While the generation tool is comprehensive, a few additional tools (e.g., for image editing or batch generation) could better match the scope.
The server's stated purpose is generating images for blogs and social media, and listing formats. The generation tool supports a wide variety of formats and options, and the format listing tool provides necessary metadata. No critical gaps are apparent.
Generate images, videos, and screenshots from your Bannerbear templates via AI.
Generate logos, social posts, app screenshots, comic panels & visual-novel assets from prompts.
AI image generation across 5 quality tiers (SDXL to Gemini 3 Pro), 50 free credits on signup.
Turn any idea, URL, doc, or PDF into on-brand visuals: 100+ formats, native in 150+ languages
