# Edytro [Health: Active]

**Category:** 🎥 Multimedia Process  
**Repository:** https://github.com/Anlans/edytro-mcp  
**GitHub Stars:** 0  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/edytro

## Description
AI image editor: remove backgrounds, upscale, and edit photos by describing the change.

## Tools
Capabilities this server exposes over MCP:

- **upload_image** — Store an image, get back a public URL
- **edit_image** — Edit or generate from a text description
- **remove_background** — Cut out a clean / transparent background
- **upscale_image** — Higher-resolution, higher-quality version
- **get_task** — Poll a task until it finishes

## Claude Desktop Quick Installation
No confirmed install details yet, so no config is published. Follow the setup steps in the repository above.

## Documentation

## What Edytro MCP server does

Edytro MCP server exposes Edytro’s AI image editing functions to MCP clients. It supports image uploads, natural-language edits, background removal, upscaling, and image generation. The service is intended for workflows where an agent needs to change an image based on a written instruction, such as replacing an object, changing clothing, or modifying a background.

The server fits the 🎥 Multimedia Process category because its tools operate on image assets and produce edited or generated image outputs. Uploaded images receive public URLs, while completed processing tasks return result URLs hosted by Edytro’s CDN.

## How it works

Image operations run asynchronously. A processing request returns a task identifier rather than waiting for the edited image to be produced. The client can then call `get_task` with that identifier to check progress and retrieve the finished image URL.

The workflow typically starts with `upload_image` when an existing image must be stored for processing. The agent can then call `edit_image`, `remove_background`, or `upscale_image`, depending on the requested transformation. `edit_image` also supports generation from a text description, so an input image is not always required for that operation.

Edytro MCP server scopes tool calls to the authenticated user’s account and available credits. Failed tasks are refunded automatically according to the README.

## Setup and configuration

The remote server is available at `https://edytro.com/mcp`. Direct connections use OAuth authentication. Edytro documents OAuth 2.1 with PKCE as the supported authentication method and also provides per-user API keys for personal use.

For ChatGPT, Claude, Cursor, or VS Code, configure the server URL and select OAuth as the authentication method. Codex users can add the `Anlans/edytro-mcp` marketplace, install Edytro from the Plugins Directory, and authorize through OAuth. The README does not specify required environment variables for the remote connection.

## Tools and capabilities

- `upload_image`: Stores an image and returns a public URL. This operation is free.
- `edit_image`: Edits an image or generates one from a text description. The listed credit cost ranges from 6 to 32 credits.
- `remove_background`: Produces an image with a clean or transparent background. The listed cost is 3 credits.
- `upscale_image`: Creates a higher-resolution, higher-quality version of an image. The listed cost is 3 credits.
- `get_task`: Checks an asynchronous operation and returns the completed image URL. This operation is free.

## Limitations and notes

Edytro MCP server requires authentication, and each tool call uses the connected user’s account and credits. Editing, background removal, and upscaling do not return final results immediately; clients must track the returned task ID with `get_task`.

The README identifies the server as a remote service rather than a locally run process. It documents OAuth and personal API keys but does not describe local installation, input size limits, supported file formats, task retention, or processing-time guarantees. Credit usage varies by operation, with `edit_image` carrying the widest documented range. Consult Edytro’s pricing, privacy, and terms pages before using it in a production workflow.

_Full upstream README: https://allmcps.com/mcp/edytro/readme_