LABSCO SUMMARY

Compare providers on the same prompt without rewiring anything.

Because the enable flags are independent, you can start with the one account you already have and add the others later, and list_models reports what is actually reachable rather than what the documentation advertises. The repository's own example runs a single architecture diagram through three providers side by side.

What it is

An MCP server that puts several image-generation providers behind one set of tools: Amazon Nova Canvas on Bedrock, OpenAI's GPT Image and DALL-E, and Google's Gemini and Imagen models.

What you get

The models currently reachable across every provider you have enabled, in one list (list_models)
An image generated from a text prompt on a model you name (generate_image)
An existing image on disk transformed according to a prompt (transform_image)
Prompt-writing guidance and worked examples, so the prompt is not the weak link (prompt_guide)
Providers enabled independently, so the server runs with only the accounts you have

Requirements

A recent Python, and at least one provider turned on: ENABLE_AWS with AWS_PROFILE and AWS_REGION, ENABLE_OPENAI with OPENAI_API_KEY, or ENABLE_GEMINI with GEMINI_API_KEY.

Setup effort

One command plus credentials — uvx universal-image-mcp@latest, then supply credentials

See installation options

USE CASES

01

Choosing which model suits a style

See what is reachable, then run the same prompt on more than one.

list_modelsgenerate_image

02

Making an illustration or diagram for a document

Write the prompt against the guidance and generate on the model you want.

prompt_guidegenerate_image

03

Reworking an image you already have

Point at the file on disk and describe the change.

transform_image

04

Getting a usable result out of a vague idea

Read the prompt guidance first, then generate.

prompt_guidegenerate_image

05

Running with only the cloud account you already pay for

Enable that provider alone and check what models it exposes.

list_models

4 tools

Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.

`generate_image`

Generate an image from a text prompt using the specified model.

Args: prompt: Detailed text description of the image to generate. Be specific about subject, style, lighting, colors, composition, and mood. Example: "A fluffy orange cat sitting on a windowsill, golden hour lighting, watercolor style" model_id: Model identifier from list_models(). Examples: "amazon.nova-canvas-v1:0", "gpt-image-1.5", "models/gemini-2.5-flash-image" output_path: Absolute or relative file path where the generated image will be saved. Supports PNG, JPEG formats. Parent directories are created automatically. reference_image: Optional. Path to an existing image to use as style/content reference. The model will generate a new image influenced by this reference. width: Optional. Image width in pixels. Default: 1024. Common values: 512, 768, 1024, 1280. Note: Some models only support specific sizes. Max: 4096px. height: Optional. Image height in pixels. Default: 1024. Common values: 512, 768, 1024, 1280. Note: Some models only support specific sizes. Max: 4096px.

Returns: Success message with output path, or error description.

BEFORE YOU INSTALL

What to know first

Tools you get

4 tools when we connected. All four tools register, but ENABLE_AWS, ENABLE_OPENAI and ENABLE_GEMINI decide which models list_models returns and which image calls succeed; with no provider enabled, nothing works.

Requires

uvx on PATH (Python 3.11+)
At least one provider: an OpenAI API key (ENABLE_OPENAI), a Gemini API key (ENABLE_GEMINI), or AWS credentials with Bedrock access (ENABLE_AWS)

Where it runs

Both documented clients run the server on your machine.

CONTEXT COST

Light Standing cost

Rated by the size of the 4 tool definitions that ride along on every turn, not by how many there are

INSTALLATION

Connect your client

Server setup

TERMINAL

uvx universal-image-mcp@latest

MCP Clients

Claude Desktop

CLAUDE DESKTOP · JSON

{
  "mcpServers": {
    "universal-image-mcp": {
      "command": "uvx",
      "args": [
        "universal-image-mcp@latest"
      ],
      "env": {
        "ENABLE_AWS": "true",
        "AWS_PROFILE": "default",
        "AWS_REGION": "us-east-1",
        "ENABLE_OPENAI": "true",
        "OPENAI_API_KEY": "<YOUR_OPENAI_API_KEY>",
        "ENABLE_GEMINI": "true",
        "GEMINI_API_KEY": "<YOUR_GEMINI_API_KEY>"
      }
    }
  }
}

Turn on only the providers you have and delete the other lines: ENABLE_OPENAI with OPENAI_API_KEY (from platform.openai.com), ENABLE_GEMINI with GEMINI_API_KEY (from Google AI Studio), ENABLE_AWS with AWS_PROFILE and AWS_REGION — the profile's keys come from aws configure or ~/.aws/credentials, and the IAM user needs Bedrock access.

Kiro

KIRO · JSON

{
  "mcpServers": {
    "universal-image-mcp": {
      "command": "uvx",
      "args": [
        "universal-image-mcp@latest"
      ],
      "env": {
        "ENABLE_AWS": "true",
        "ENABLE_OPENAI": "true",
        "OPENAI_API_KEY": "<YOUR_OPENAI_API_KEY>",
        "ENABLE_GEMINI": "true",
        "GEMINI_API_KEY": "<YOUR_GEMINI_API_KEY>"
      }
    }
  }
}

Save as ~/.kiro/settings/mcp.json. Turn on only the providers you have: ENABLE_OPENAI with OPENAI_API_KEY, ENABLE_GEMINI with GEMINI_API_KEY; ENABLE_AWS uses the AWS credentials already on your machine (aws configure or ~/.aws/credentials), or add AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY and AWS_REGION to env.

Universal Image MCP

Compare providers on the same prompt without rewiring anything.

What it is

What you get

Requirements

Setup effort

Choosing which model suits a style

Making an illustration or diagram for a document

Reworking an image you already have

Getting a usable result out of a vague idea

Running with only the cloud account you already pay for

4 tools

`generate_image`

`list_models`

`prompt_guide`

`transform_image`

What to know first

Tools you get

Requires

Where it runs

Light Standing cost

Connect your client

Server setup

MCP Clients

Claude Desktop

Kiro

Universal Image MCP

Compare providers on the same prompt without rewiring anything.

What it is

What you get

Requirements

Setup effort

Choosing which model suits a style

Making an illustration or diagram for a document

Reworking an image you already have

Getting a usable result out of a vague idea

Running with only the cloud account you already pay for

4 tools

generate_image

list_models

prompt_guide

transform_image

What to know first

Tools you get

Requires

Where it runs

Light Standing cost

Connect your client

Server setup

MCP Clients

Claude Desktop

Kiro

Same category

OpenAI GPT Image

Nano Banana MCP

Imagen3-MCP

Gemini Imagen 3.0

`generate_image`

`list_models`

`prompt_guide`

`transform_image`