Get the FREE Ultimate OpenClaw Setup Guide →

qwen-image-replicate

A Model Context Protocol (MCP) server that provides image generation capabilities using Alibaba's Qwen Image model via the Replicate platform.

Installation
Run this command in your terminal to add the MCP server to Claude Code.
Run in terminal:
Command
claude mcp add --transport stdio pierrunoyt-qwen-image-replicate-mcp-server npx -y https://github.com/PierrunoYT/qwen-image-fal-mcp-server.git \
  --env REPLICATE_API_TOKEN="your_replicate_token_here"

How to use

The Qwen Image Replicate MCP Server exposes image generation capabilities by leveraging Alibaba's Qwen Image model through the Replicate platform. The server provides a generate_image tool that accepts prompts and several optional parameters to control image size, quality, and style. Typical usage involves sending a prompt and choosing options such as image_size, number of inference steps, CFG guidance scale, and a seed for reproducibility. You can wire this server into your MCP client (Claude Desktop, Kilo Code, or any MCP-enabled client) and issue the generate_image capability to obtain generated images via Replicate.

When using the server, specify prompts that describe the scene or concept you want, along with any desired image characteristics. The available parameters include: prompt (required), image_size (optional with defaults), num_inference_steps (optional, 1-500), seed (optional), guidance_scale (optional, 1-20), and negative_prompt (optional). The server handles the Replicate API interaction, including token authentication via the REPLICATE_API_TOKEN environment variable. If your environment lacks the API key, the server will still run but will skip actual generation until a valid key is provided.

How to install

Prerequisites:\n- Node.js v18+ installed on your machine.\n- A Replicate API key.\n- Access to the internet to fetch the MCP server from GitHub.\n\nQuick Setup with npx (recommended):\n1. Ensure Node.js is installed and you have a Replicate API key.\n2. In your MCP client configuration, add the qwen-image server using npx as shown in the example. The setup will download and run the latest version directly from GitHub.\n\nManual Installation (alternative):\n1. Clone the repository and install dependencies locally:\nbash\ngit clone https://github.com/PierrunoYT/qwen-image-fal-mcp-server.git\ncd qwen-image-fal-mcp-server\nnpm install\n 2. Build the server (if applicable):\nbash\nnpm run build\n 3. Run the server locally (example path to built entry):\nbash\nnode build/index.js\n 4. In your MCP configuration, reference the absolute path to the server entry (see README examples) or rely on npx for automatic retrieval. Ensure the environment variable REPLICATE_API_TOKEN is set for authentication.

Additional notes

Environment variables and configuration tips:\n- REQUIRED: REPLICATE_API_TOKEN must be set to your Replicate API key to perform image generation.\n- Optional: NODE_ENV, LOG_LEVEL, MAX_CONCURRENT_REQUESTS, REQUEST_TIMEOUT can be tuned for performance and debugging.\n- If you encounter issues with missing API token, the server will continue to run post-setup but generation requests will fail until the key is supplied.\n- When using the npx setup, you automatically get the latest version from GitHub. If you prefer a local install, ensure you keep dependencies up to date and point MCP to the correct absolute path to the built entry file.\n- For best results, consider using a reasonable default image_size (e.g., 1024x768) and experiment with num_inference_steps and guidance_scale to balance quality and speed.

Related MCP Servers

Sponsor this space

Reach thousands of developers