What it does
This remote MCP server exposes an AI image editor that edits photos from natural-language instructions. Its tools upload an image and return a public URL, edit or generate images from a text description, remove backgrounds to a clean or transparent cut-out, and upscale images to a higher resolution and quality. Operations run asynchronously: a create call returns a task_id straight away, and get_task returns the finished image URL, hosted on a CDN. Failed tasks are refunded automatically.
It suits situations where an assistant needs to change photos through plain instructions, such as replacing or removing backgrounds, changing objects or clothing, upscaling an image, or generating a new one. It can be connected from ChatGPT, Codex, Claude, Cursor and VS Code.
To run it, connect the server to a supported client. OAuth 2.1 with PKCE is the supported authentication method, and per-user API keys are available for personal use. Every tool call is scoped to the connected user's own account and credits, and the create and polling flow means results are retrieved in a separate step.
Limitations and requirements
Tool calls use credits from the connected user's account: edit_image costs 6 to 32 credits, remove_background and upscale_image cost 3 credits each, while upload_image and get_task are free. Access requires authentication through OAuth 2.1 or a per-user API key, and calls are limited to the connected user's own account and credits.
Safety
Our team has not reviewed this resource. Only automated rules checked it; resources with high-risk signals are not published. These are signals, not a guarantee that installing it is safe. Try it in a sandbox project first.
- Runs code or changes an assistant's behaviour
Sources
- Official MCP Registry first seen 13 Nov 2026 · last seen 13 Nov 2026 · its page on the source
Edytro
AI image editor: remove backgrounds, upscale, and edit photos by describing the change.
v1.0.0As shown on registry.modelcontextprotocol.io when we last read it