FGCLIP-MCP
MCP (Model Context Protocol) server for FG-CLIP embedding services. To obtain and configure the API key, please apply at https://research.360.cn/sass.
Features
This MCP server provides the following tools and resources:
Power your agent
Install from the publisher. Not a PlugMCP endpoint.
Host a copy with our keys and logs.
No packaged release in the registry yet. Follow the repository README.
Remote URL or a command on your machine? How the two installs differ
Run on PlugMCP
A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.
LaunchFrom the publisher
MCP (Model Context Protocol) server for FG-CLIP embedding services. To obtain and configure the API key, please apply at https://research.360.cn/sass.
This MCP server provides the following tools and resources:
MCP (Model Context Protocol) server for FG-CLIP embedding services.
This MCP server helps users achieve the following capabilities:
Through these capabilities, users can build powerful search engines, recommendation systems, content classification, and multi-modal AI applications.
Generate embedding vectors for input texts.
Parameters:
texts: A list of text strings to embedmodel: The model to use (default: "fg-clip")Returns:
saved_uris: A list of URIs where the embeddings are storedsuccess: Whether the operation succeedederror_msg: Error message, if anyGenerate embedding vectors for images.
Parameters:
images: A list of image URLs or base64-encoded imagesmodel: The model to use (default: "fg-clip")Returns:
saved_uris: A list of URIs where the embeddings are storedsuccess: Whether the operation succeedederror_msg: Error message, if anyCompute cosine similarity between two lists of vectors.
Parameters:
uris_a: A list of URIs for the first set of embeddingsuris_b: A list of URIs for the second set of embeddingsmode: Calculation mode (default: "pairwise")
"pairwise": Compute similarity for vectors at corresponding positions"matrix": Compute a full similarity matrix for all vector pairsReturns:
similarities: Similarity values or a similarity matrixshape: Shape information of the resultsuccess: Whether the operation succeededgit clone https://github.com/360CVGroup/FGCLIP-MCP
cd FGCLIP-MCP
uv venv
uv sync
source .venv/bin/activate
export MCP_API_KEY=your_api_key
pytest -q
{
"mcpServers": {
"fgclip-mcp": {
"command": "uvx",
"args": [
"fgclip-mcp"
],
"env": {
"MCP_API_KEY": "your_api_key"
}
}
}
}
{
"mcpServers": {
"fgclip-mcp-local": {
"command": "uv",
"args": [
"--directory",
"/path_to_fgclip-mcp/src/fgclip_mcp",
"run",
"/path_to_fgclip-mcp/src/fgclip_mcp/__main__.py"
],
"env": {
"MCP_API_KEY": "your_api_key"
}
}
}
}
Locate MCP Setting

Config MCP Setting

Enable MCP

Chat with MCP
Example: Searching for images based on given text

Image URLs:
Apache License 2.0
MCP (Model Context Protocol) server for FG-CLIP embedding services.
Tools
No tool list yet. Remote servers get one after the first health check.
FAQ
Copy the stdio config for Cursor, Claude Desktop, Claude Code, or VS Code on this page. It uses the publisher's install, not a PlugMCP proxy.
No required environment variable is listed on this record.
FGCLIP runs locally over stdio.
Last checked 2026-10-26. MCP (Model Context Protocol) server for FG-CLIP embedding services.