unsplash-mcp-server

MCP.Pizza Chef: hellokaton

Unsplash MCP Server is a lightweight MCP server that integrates Unsplash's extensive photo library into applications. It provides advanced image search capabilities including keyword relevance, color filtering, orientation options, custom sorting, and pagination. Designed for developers, it enables seamless access to high-quality images via the Unsplash API, enhancing apps with rich visual content.

Use This MCP server To

Search Unsplash images by keyword relevance Filter images by color schemes Filter images by orientation (landscape, portrait) Sort image search results with custom criteria Paginate through large image search results Integrate high-quality Unsplash photos into apps Enable dynamic image retrieval for content generation Support visual content discovery in chatbots or AI agents

README

Unsplash MCP Server

English | 简体中文

A simple MCP server for seamless Unsplash image integration and search capabilities.

Python 3.9+ License: MIT smithery badge

📋 Overview

Unsplash MCP Server is used for searching rich, high-quality images. It's ideal for developers who want to integrate Unsplash functionality into their own applications.

✨ Features

  • Advanced Image Search: Search Unsplash's extensive photo library with filters for:
    • Keyword relevance
    • Color schemes
    • Orientation options
    • Custom sorting and pagination

🔑 Obtaining Unsplash Access Key

Before installing this server, you'll need to obtain an Unsplash API Access Key:

  1. Create a developer account at Unsplash
  2. Register a new application
  3. Get your Access Key from the application details page
  4. Use this key in the configuration steps below

For more details, refer to the official Unsplash API documentation.

🚀 Installation

To install Unsplash Image Integration Server for Claude Desktop automatically via Smithery:

IDE Setup

Cursor IDE

npx -y @smithery/cli@latest install @hellokaton/unsplash-mcp-server --client cursor --key 7558c683-****-****

Windsurf

npx -y @smithery/cli@latest install @hellokaton/unsplash-mcp-server --client windsurf --key 7558c683-****-****

Cline

npx -y @smithery/cli@latest install @hellokaton/unsplash-mcp-server --client cline --key 7558c683-****-****

Manual Installation

# Clone the repository
git clone https://github.com/hellokaton/unsplash-mcp-server.git

# Navigate to project directory
cd unsplash-mcp-server

# Create virtual environment
uv venv

# Install dependencies
uv pip install .

Cursor Editor Integration

Add the following configuration to your Cursor editor's settings.json:

⚠️ Note: Please adjust the following configuration according to your actual installation:

  • If uv is not in your system PATH, use an absolute path (e.g., /path/to/uv)
  • ./server.py should be modified to the actual location of your server script (can use absolute path or path relative to workspace)

Cursor Configuration Screenshot

{
  "mcpServers": {
    "unsplash": {
      "command": "uv",
      "args": ["run", "--with", "fastmcp", "fastmcp", "run", "./server.py"],
      "env": {
        "UNSPLASH_ACCESS_KEY": "${YOUR_ACCESS_KEY}"
      }
    }
  }
}

Using in Cursor

Unsplash MCP in Cursor

🛠️ Available Tools

Search Photos

{
  "tool": "search_photos",
  "query": "mountain",
  "per_page": 5,
  "orientation": "landscape"
}

🔄 Other Implementations

📄 License

MIT License

📬 Contact

unsplash-mcp-server FAQ

How do I obtain an Unsplash API Access Key?
You must register on Unsplash Developers portal to get an API Access Key required for this server.
What programming language is the Unsplash MCP Server built with?
It is built using Python 3.9+ for easy integration and deployment.
Can I filter image searches by color or orientation?
Yes, the server supports filtering by color schemes and image orientation such as landscape or portrait.
Does the server support pagination for large result sets?
Yes, it supports custom pagination to navigate through extensive image search results.
Is the Unsplash MCP Server open source?
Yes, it is licensed under the MIT License and available on GitHub.
Can this server be used with multiple LLM providers?
Yes, it is provider-agnostic and can be integrated with models like OpenAI, Claude, and Gemini.