Grok Image
MCP server giving AI clients image generation, editing and headshot tools via xAI Grok
- Pricing
- Open source
- Platforms
- CLI, API, self-hosted
- Version
- 0.1.0
- Listing updated
- today
Similar MCP servers
About Grok Image
Grok Image is an open-source MCP server for design and media. It is available on CLI, API and self-hosted.
The README describes this as a Rust-built MCP server that connects AI clients to xAI's Grok image API over stdio using JSON-RPC 2.0. Instead of calling the image endpoints directly, a client can invoke named tools for generating images from text, editing one or several source images with natural language, and producing padded corporate headshots.
A notable feature is its style system. Fourteen built-in styles such as watercolor, pixel-art and noir wrap a user's prompt in a template, and users can add their own styles or override built-ins through a TOML config file. Responses echo back the resolved prompt so the sent text can be checked.
The server needs an xAI API key stored in a local config file and is set up by adding its binary path to an MCP client configuration. It suits developers and people who already use an MCP-capable assistant and want image work available inside it.
Tools this server provides (4)
Taken from the project documentation; the running server may differ.
- generate_image
- Generate an image from a text prompt
- edit_image
- Edit an existing image using natural language instructions
- headshot
- Corporate headshot from a source portrait
- list_styles
- List available image styles for use with generate_image
What it does
- Text-to-image generation
- Creates images from a prompt with control over aspect ratio, resolution, quality and count.
- Multi-image editing
- Edits one image or up to five reference images using natural language instructions.
- Style templates
- Applies fourteen built-in styles or custom templates to generation prompts.
- Headshot preparation
- Expands portraits to a wider canvas with white gutters without cropping the subject.
- Flexible output
- Returns temporary URLs or base64 JSON data for each image.
- Local config
- Reads the API key and custom styles from a TOML file in the user's config directory.
Good for
- Generate illustrations from text inside an MCP-enabled AI assistant
- Edit a product or scene photo with instructions like removing the background
- Produce consistent artwork by applying a named style to multiple prompts
- Prepare a corporate headshot with a clean white background from an existing portrait
Pricing
Open source
MIT-licensed. Image generation is billed by xAI against your own API key.
Grok Image FAQ
Which tools does the server expose?
It exposes generate_image, edit_image, headshot and list_styles.
Does edit_image support a style?
No. The README says the style parameter is intentionally left off edit_image because edit prompts are instructions rather than descriptions.
How many images can be edited at once?
Up to five source images can be referenced in one edit request using the grok-imagine-image-2.0 model.
What do I need to run it?
Rust to build it, and an xAI API key placed in a config file.
Is Grok Image free?
Yes. Grok Image is open source, so you can use and self-host it for free.
What are the best alternatives to Grok Image?
The closest alternatives to Grok Image are Grok Chat, Muse AI Image MCP, AI Image Editor MCP, AI Photo Editor MCP and Best AI Image Prompt MCP.
Where this comes from
- Description: written from the project's README and its MCP Registry entry.
