What can you do with this server?

Grok MCP is an MCP server that provides comprehensive access to xAI's Grok API capabilities. You can access multiple Grok models (Grok-4, Grok-4-Fast, Grok-3-Mini, and more) for chat completion with extensive customization options including temperature, max tokens, and system prompts. The server supports reasoning models that provide detailed reasoning alongside responses, image generation from text descriptions, and vision analysis of images using natural language queries (supporting both local files and URLs). It offers live web search with real-time results, source citations, date range filters, country localization, and custom RSS feed integration from news, web, X (Twitter), and RSS sources. You can maintain stateful conversations with context preserved across multiple requests, conversation history management, and the ability to retrieve and delete stored responses (kept for 30 days). Additional features include model discovery to list all available Grok models with their details and creation information.

Which integrations are available for this server?

Implemented in Python and requires Python 3.11+ runtime environment for execution Enables real-time web search with RSS feed integration, allowing searches across custom RSS sources with citation support

How do I use Grok MCP?

1. Click on "Install Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Grok MCP generate an image of a futuristic city skyline at sunset" 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.

Grok-MCP

MCP server for xAI's Grok API with agentic tool calling, image generation, vision, and reasoning models.

Features

Agentic Tool Calling: Web search, X search, and code execution with multi-step reasoning
Multiple Grok Models: Access to Grok-4.1-Fast-Reasoning, Grok-4.1-Fast-Non-Reasoning, Grok-4-Fast, Grok-3-Mini, and more
Image Generation: Create images using Grok's image generation model
Vision Capabilities: Analyze images with Grok's vision models
Reasoning Models: Advanced reasoning with extended thinking models (Grok-4.1-Fast-Reasoning, Grok-3-Mini, Grok-4)
Stateful Conversations: Use this newly released feature to maintain conversation context as id across multiple requests

Related MCP server: Zen MCP Server

Prerequisites

Python 3.11 or higher
xAI API key (Get one here)
Astral UV

Installation

Clone the repository:

git clone https://github.com/merterbak/Grok-MCP.git cd Grok-MCP

Create a venv environment:

uv venv source .venv/bin/activate # macOS/Linux or .venv\Scripts\activate on Windows

Install dependencies:

uv sync

Configuration

Claude Desktop Integration

Add this to your Claude Desktop configuration file:

{ "mcpServers": { "grok": { "command": "uv", "args": [ "--directory", "/path/to/Grok-MCP", "run", "python", "main.py" ], "env": { "XAI_API_KEY": "your_api_key_here" } } } }

Filesystem MCP (Optional)

Claude Desktop can't send uploaded images in the chat to an MCP tool. The easiest way to give access to files directly from your computer is official Filesystem MCP server. After setting it up you’ll be able to just write the image’s file path (such as /Users/mert/Desktop/image.png) in chat and Claude can use it with any vision chat tool.

{ "mcpServers": { "filesystem": { "command": "npx", "args": [ "-y", "@modelcontextprotocol/server-filesystem", "/Users/<your-username>/Desktop", "/Users/<your-username>/Downloads" ] } } }

For stdio:

uv run python main.py

Docker:

docker compose up --build

Mcp Inspector:

mcp dev main.py

Available Tools

Note: For using images and files, you must provide paths to chat. See Filesystem MCP (Optional) for setup.

`list_models`

List all available Grok models.

`chat`

Standard chat completion.

Parameter	Type	Default	Description
`prompt`	str	required	Your message
`model`	str	grok-4	Model to use
`system_prompt`	str	None	System instruction
`store_messages`	bool	False	Enable conversation history

`chat_with_vision`

Analyze images with text.

Parameter	Type	Default	Description
`prompt`	str	required	Question about the image
`image_paths`	List[str]	None	Local image file paths
`image_urls`	List[str]	None	Image URLs
`detail`	str	auto	auto, low, or high
`model`	str	grok-4	Vision model

Returns: Content + usage with prompt_image_tokens

`chat_with_reasoning`

Get detailed reasoning with the response.

Parameter	Type	Default	Description
`prompt`	str	required	Your question
`model`	str	grok-3-mini	Reasoning model
`reasoning_effort`	str	None	low or high

Returns: Content, reasoning_content, usage (with reasoning_tokens)

`generate_image`

Create images from text.

Parameter	Type	Default	Description
`prompt`	str	required	Image description
`n`	int	1	Number of images
`image_format`	str	url	url or b64_json
`model`	str	grok-2-image-1212	Image model

`web_search`

Agentic web search with autonomous research.

Parameter	Type	Default	Description
`prompt`	str	required	Search query
`model`	str	grok-4-1-fast	Model
`allowed_domains`	List[str]	None	Restrict to domains (max 5)
`excluded_domains`	List[str]	None	Exclude domains (max 5)
`enable_image_understanding`	bool	False	Analyze images in results
`include_inline_citations`	bool	False	Embed citations in text
`max_turns`	int	None	Limit reasoning turns

Returns: Content, citations, tool_calls, usage

`x_search`

Agentic X (Twitter) search.

Parameter	Type	Default	Description
`prompt`	str	required	Search query
`model`	str	grok-4-1-fast	Model
`allowed_x_handles`	List[str]	None	Only these handles (max 10)
`excluded_x_handles`	List[str]	None	Exclude handles (max 10)
`from_date`	str	None	Start date (DD-MM-YYYY)
`to_date`	str	None	End date (DD-MM-YYYY)
`enable_image_understanding`	bool	False	Analyze images
`enable_video_understanding`	bool	False	Analyze videos
`include_inline_citations`	bool	False	Embed citations
`max_turns`	int	None	Limit turns

Returns: Content, citations, tool_calls, usage

`grok_agent`

Unified agent combining files, images, and all agentic tools (web search, X search, code execution).

Parameter	Type	Default	Description
`prompt`	str	required	Your query
`file_ids`	List[str]	None	Uploaded file IDs to search
`image_urls`	List[str]	None	Image URLs to analyze
`image_paths`	List[str]	None	Local image paths
`use_web_search`	bool	False	Enable web search
`use_x_search`	bool	False	Enable X search
`use_code_execution`	bool	False	Enable code execution
+ all web_search and x_search params

Returns: Content, citations, tool_calls, code_outputs, uploaded_file_ids, usage

`code_executor`

Execute Python code for calculations and analysis.

Parameter	Type	Default	Description
`prompt`	str	required	Task description
`model`	str	grok-4-1-fast	Model
`max_turns`	int	None	Limit turns

Returns: Content, tool_calls, code_outputs, usage

`stateful_chat`

Maintain conversation state across requests.

Parameter	Type	Default	Description
`prompt`	str	required	Your message
`response_id`	str	None	Previous response ID
`model`	str	grok-4	Model
`system_prompt`	str	None	System instruction

Returns: Content, response_id, usage

`retrieve_stateful_response`

Retrieve a stored conversation.

`delete_stateful_response`

Delete a stored conversation.

`upload_file`

Upload a document (max 48 MB).

Parameter	Type	Default	Description
`file_path`	str	required	Local file path

Supported formats: .txt, .md, .py, .js, .csv, .json, .pdf, and more

`list_files`

List uploaded files with sorting.

Parameter	Type	Default	Description
`limit`	int	100	Max files to return
`order`	str	desc	asc or desc
`sort_by`	str	created_at	created_at, filename, or size

`get_file`

Get file metadata by ID.

`get_file_content`

Download file content by ID.

Parameter	Type	Default	Description
`file_id`	str	required	File ID
`max_bytes`	int	500000	Max bytes to return

`delete_file`

Delete a file by ID.

`chat_with_files`

Chat with uploaded documents using agentic document search.

Parameter	Type	Default	Description
`prompt`	str	required	Question about docs
`file_ids`	List[str]	required	File IDs to search
`model`	str	grok-4-1-fast	Model
`system_prompt`	str	None	System instruction

Returns: Content, citations, usage

License

This project is open source and available under the MIT License.

Install Server

A

security – no known vulnerabilities

A

license - permissive license

A

quality - confirmed to work

How are these scores calculated?

Resources

GitHub Repository

Need Help?

Report Issue

Related Servers

Tools

View all tools

Grok MCP