
Security News
Deno 2.4 Brings Back deno bundle, Improves Dependency Management and Observability
Deno 2.4 brings back bundling, improves dependency updates and telemetry, and makes the runtime more practical for real-world JavaScript projects.
Session-based conversational image generation MCP server using OpenAI's advanced models
A Model Context Protocol (MCP) server that provides conversational OpenAI image generation capabilities. Generate, edit, and refine images through multi-turn conversations with advanced models like GPT-4o and GPT-4.1.
ā Single-shot limitations - "Make it more blue" requires re-describing everything
ā No conversation memory - Each request starts from scratch
ā Context loss - Can't reference previous images naturally
ā Manual workflows - Complex multi-step processes require multiple tools
ā
Conversational refinement - "Make it more blue" works naturally
ā
Session memory - Builds on previous context automatically
ā
Reference awareness - "Use the same style as the previous image"
ā
Integrated workflows - Single interface for complex creative projects
# Start a focused session
session = create_image_session("Logo design for tech startup")
# Initial generation
result1 = generate_image_in_session(session_id, "modern tech logo")
# Natural refinement - no need to repeat everything
result2 = generate_image_in_session(session_id, "make it more minimalist")
# Build on context
result3 = generate_image_in_session(session_id, "try it in dark blue")
Start simple, expand when needed:
# Quick one-shot for immediate need
result = generate_image("modern office workspace")
# Later, promote to session for refinement
session = promote_image_to_session(
result["image_path"],
"Office workspace refinement project"
)
# Continue with conversational context
generate_image_in_session(session_id, "add more plants and warmer lighting")
Requirements: Python 3.10 or higher
# Install the package
pip install openai-image-mcp
If you need to upgrade Python:
# Using pyenv (recommended)
pyenv install 3.11.8
pyenv global 3.11.8
pip install openai-image-mcp
# Or using Homebrew (macOS)
brew install python@3.11
python3.11 -m pip install openai-image-mcp
For development installation from source, see DEVELOPMENT.md
Add to your Claude Desktop MCP configuration:
{
"mcpServers": {
"openai-image-mcp": {
"command": "sh",
"args": [
"-c",
"openai-image-mcp 2> mcp_server_stderr.log"
],
"env": {
"OPENAI_API_KEY": "your_openai_api_key_here"
}
}
}
}
For development setup and alternative configurations, see DEVELOPMENT.md
# Create a session for your project
session = create_image_session("Website hero images")
# Generate with natural language
generate_image_in_session(session_id, "modern tech office with diverse team")
# Refine naturally
generate_image_in_session(session_id, "make the lighting warmer")
# Add context
generate_image_in_session(session_id, "create a mobile version of this scene")
create_image_session
- Start conversational sessiongenerate_image_in_session
- Generate with context awarenessget_session_status
- View conversation history and progressclose_session
- End session and cleanupgenerate_image
- General purpose (session optional)edit_image
- Modify existing imagesgenerate_product_image
- E-commerce optimizedgenerate_ui_asset
- UI/UX design elementsanalyze_and_improve_image
- AI-powered image enhancementpromote_image_to_session
- Upgrade one-shot to conversationallist_active_sessions
- Manage multiple projectsget_usage_guide
- Comprehensive tool documentationBest for: Multi-image projects, iterative refinement, brand consistency
session = create_image_session("App icon design")
generate_image_in_session(session_id, "colorful chat app icon")
generate_image_in_session(session_id, "make it more professional")
generate_image_in_session(session_id, "try different color schemes")
Best for: Immediate needs, single images, uncertain scope
generate_image("professional headshot for LinkedIn")
generate_product_image("wireless headphones", background_type="white")
Best for: Testing concepts, uncertain requirements, flexible workflows
# Start quick
result = generate_image("logo concept for bakery")
# Expand when needed
session = promote_image_to_session(result["image_path"], "Bakery brand development")
generate_image_in_session(session_id, "create business card version")
session = create_image_session("TechCorp brand identity")
# Logo concepts
generate_image_in_session(session_id, "modern tech company logo")
generate_image_in_session(session_id, "make it more geometric and minimal")
# Expand to brand elements
generate_image_in_session(session_id, "business card design using this logo")
generate_image_in_session(session_id, "website header with the logo")
session = create_image_session("Wireless headphones marketing")
# Product shots
generate_product_image("premium wireless headphones", angle="45deg")
result = promote_image_to_session(previous_result["image_path"], "headphones campaign")
# Marketing variations
generate_image_in_session(session_id, "lifestyle shot with person using them")
generate_image_in_session(session_id, "create packaging design mockup")
OPENAI_API_KEY
(required) - Your OpenAI API keyMCP_MAX_SESSIONS
(optional) - Maximum concurrent sessions (default: 100)MCP_SESSION_TIMEOUT
(optional) - Session timeout in seconds (default: 3600)We welcome contributions! Please see DEVELOPMENT.md for:
This project is licensed under the MIT License - see the LICENSE file for details.
FAQs
Session-based conversational image generation MCP server using OpenAI's advanced models
We found that openai-image-mcp demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago.Ā It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Security News
Deno 2.4 brings back bundling, improves dependency updates and telemetry, and makes the runtime more practical for real-world JavaScript projects.
Security News
CVEForecast.org uses machine learning to project a record-breaking surge in vulnerability disclosures in 2025.
Security News
Browserslist-rs now uses static data to reduce binary size by over 1MB, improving memory use and performance for Rust-based frontend tools.