Mcp Vision

What is Mcp Vision about?

Mcp Vision provides a Model Context Protocol (MCP) server that wraps HuggingFace computer‑vision models—primarily zero‑shot object detection pipelines—into callable tools. These tools can be invoked by LLMs (e.g., Claude) to locate objects in an image or to crop and zoom into a specific object.

How to use Mcp Vision?

1. Installation
   • Clone the repository.
   • Build the Docker image with make build-docker (or use the pre‑built image groundlight/mcp-vision:latest).
2. Configure your client (e.g., Claude Desktop) by adding an entry to claude_desktop_config.json that runs the Docker container, choosing the GPU or CPU variant.
3. Call the tools from a prompt, providing an image path/URL and the desired labels or target object.

Key Features

• locate_objects: Zero‑shot detection of any set of candidate labels; returns detections in HuggingFace format.
• zoom_to_object: Automatically crops the highest‑scoring detection of a specified label and returns the cropped image.
• Supports GPU acceleration via NVIDIA runtime; falls back to CPU with optional smaller default model.
• Easy Docker‑based deployment; also runnable locally via the uv Python package manager.

Use Cases

• Enhancing LLM‑driven agents with visual reasoning (e.g., answering questions about storefront signage).
• Pre‑processing images for downstream vision‑language pipelines by extracting regions of interest.
• Rapid prototyping of custom object‑detection‑as‑a‑service without training new models.

FAQ

Q: Do I need a GPU? A: A GPU speeds up inference dramatically. The default google/owlvit-large-patch14 model is slow on CPU; you can set DEFAULT_OBJDET_MODEL to a smaller model.

Q: Can I run the server without building the Docker image? A: Yes, pull groundlight/mcp-vision:latest and run it directly, though the initial model download may delay startup.

Q: Why does Claude sometimes ignore the tools? A: If web‑search is enabled, Claude may prefer that over local MCP tools. Disable web‑search for best results.

Q: How do I add more models or tools? A: Extend the codebase under mcp_vision/tools and update the Docker image; a TODO mentions hosting best models online to avoid local downloads.

mcp-vision by

A Model Context Protocol (MCP) server exposing HuggingFace computer vision models such as zero-shot object detection as tools, enhancing the vision capabilities of large language or vision-language models.

This repo is in active development. See below for details of currently available tools.

Installation

Clone the repo:

git clone git@github.com:groundlight/mcp-vision.git

Build a local docker image:

cd mcp-vision
make build-docker

Configuring Claude Desktop

Add this to your claude_desktop_config.json:

If your local environment has access to a NVIDIA GPU:

"mcpServers": {
  "mcp-vision": {
    "command": "docker",
    "args": ["run", "-i", "--rm", "--runtime=nvidia", "--gpus", "all", "mcp-vision"],
	"env": {}
  }
}

Or, CPU only:

"mcpServers": {
  "mcp-vision": {
    "command": "docker",
    "args": ["run", "-i", "--rm", "mcp-vision"],
	"env": {}
  }
}

When running on CPU, the default large-size object detection model make take a long time to laod and run inference. Consider using a smaller model as DEFAULT_OBJDET_MODEL (you can tell Claude directly to use a specific model too).

(Beta) It is possible to run the public docker image directly without building locally, however the download time may interfere with Claude's loading of the server.

"mcpServers": {
  "mcp-vision": {
    "command": "docker",
    "args": ["run", "-i", "--rm", "--runtime=nvidia", "--gpus", "all", "groundlight/mcp-vision:latest"],
	"env": {}
  }
}

Tools

The following tools are currently available through the mcp-vision server:

1. locate_objects

• Description: Detect and locate objects in an image using one of the zero-shot object detection pipelines available through HuggingFace (list for reference [https://huggingface.co/models?pipeline_tag=zero-shot-object-detection&sort=trending]).
• Input: image_path (string) URL or file path, candidate_labels (list of strings) list of possible objects to detect, hf_model (optional string), will use "google/owlvit-large-patch14" by default, which could be slow on a non-GPU machine
• Returns: List of dicts in HF object-detection format

2. zoom_to_object

• Description: Zoom into an object in the image, allowing you to analyze it more closely. Crop image to the object bounding box and return the cropped image. If many objects are present in the image, will return the 'best' one as represented by object score.
• Input: image_path (string) URL or file path, label (string) object label to find and zoom and crop to, hf_model (optional), will use "google/owlvit-large-patch14" by default, which could be slow on a non-GPU machine
• Returns: MCPImage or None

Example in blog post and video

Run Claude Desktop with Claude Sonnet 3.7 and mcp-vision configured as an MCP server in claude_desktop_config.json.

The prompt used in the example video and blog post was:

From the information on that advertising board, what is the type of this shop?
Options:
The shop is a yoga studio.
The shop is a cafe.
The shop is a seven-eleven.
The shop is a milk tea shop.

The image is the first image in the V*Bench/GPT4V-hard dataset and can be found here: https://huggingface.co/datasets/craigwu/vstar_bench/blob/main/GPT4V-hard/0.JPG (use the download link).

Note:

• If you upload the image directly into the conversation with Claude instead of providing a download link, it will not be able to call the tools and will attempt to answer directly.
• On accounts that have web search enabled, Claude will prefer to use web search over local MCP tools AFAIK. Disable web search for best results.

Development

Run locally using the uv package manager:

uv install
uv run python mcp_vision

Build the Docker image locally:

make build-docker

Run the Docker image locally:

make run-docker-cpu

or

make run-docker-gpu

[Groundlight Internal] Push the Docker image to Docker Hub (requires DockerHub credentials):

make push-docker

Troubleshooting

If Claude Desktop is failing to connect to mcp-vision:

• Check the configuration is correct (CPU vs GPU)
• Developer options may need to be enabled in Claude Desktop
• Depending on the size of the model(s) used, give it a few minutes to download them from HuggingFace on first opening Claude Desktop. Once downloaded, the server will respond and Claude will connect.

On accounts that have web search enabled, Claude will prefer to use web search over local MCP tools AFAIK. Disable web search for best results.

TODO

• Host best models online instead of requiring local download
• Add more tools