What the MCP server is

MCP (Model Context Protocol) lets an AI assistant such as Claude or Cursor work with your 3DPrinterOS account directly. Once it is connected, you can ask the assistant in plain words to check printers, upload files, queue or start prints, cancel jobs, share printers, pull reports, and manage organization users. The assistant picks the right 3DPrinterOS action, runs it with your account permissions, and tells you the result.

The MCP server is based on the 3DPrinterOS Global API calls v2.1, but its tools are not an exact copy of the API calls. Some tool names are different, some API calls are left out, and a few tools exist only in the MCP server. You also do not need to log in with a session or send API requests yourself.

The MCP server works for accounts with a license or an active trial.

Server URL	`https://cloud.3dprinteros.com/apimcp`
Transport	Streamable HTTP
Authentication	Header `Authorization: Bearer YOUR_TOKEN`
Number of tools	31
Request limit	120 requests per minute

Get your MCP token

Log in to cloud.3dprinteros.com with the account the AI should work as.
Open your Account page.
Scroll to the MCP Server Access token section at the bottom and click Enable.
In the MCP Token Enable window, read the Data Processing Notice and tick I understand my data could be processed externally.
Select at least one permission for the token (see the table below), then click Enable.
Copy the token and add it to your AI app as shown below.

Permission	What it allows
`read`	Read access to your own resources, such as printers, jobs, and projects.
`write`	Write access to your own resources, such as uploading files and starting or cancelling jobs.
`org:read`	Read access to organization resources, such as users, workgroups, and reports. Org admins only.
`org:write`	Write access to organization resources, such as creating users and changing balances. Org admins only.

Select only the permissions the AI needs. For example, a token with only read and org:read lets the AI check printers and pull reports without changing anything.

The token expires after 7 days, and the 7 days restart while you keep using it. If you generate a new token, the old one stops working right away, so update every AI app that used it.

The AI has your permissions. Everything the assistant does runs as your account. With an org admin token it can create users, change balances, and start prints, so treat the token like a password and do not share it or commit it to a code repository.

An org member's token works for that member's own printers, files, projects, and shared printers. Organization management tools (users, workgroups, balances, reports) return an access error for a member token.

Connect your AI app

Claude Code

Run this command in a terminal and replace YOUR_TOKEN with your token.

claude mcp add --transport http 3dprinteros https://cloud.3dprinteros.com/apimcp --header "Authorization: Bearer YOUR_TOKEN"

Add --scope user to make it available in all your projects. The default scope is the current project only. Type /mcp inside Claude Code to check that the server is connected.

To keep the token out of the command, add the server to a .mcp.json file and read the token from an environment variable.

{"mcpServers":{"3dprinteros":{"type":"http","url":"https://cloud.3dprinteros.com/apimcp","headers":{"Authorization":"Bearer ${PRINTEROS_MCP_TOKEN}"}}}}

Claude (web and desktop app)

Go to Customize > Connectors.
Click + Add, then Add custom connector.
Enter a name (for example, 3DPrinterOS) and the server URL https://cloud.3dprinteros.com/apimcp, then click Continue.
Under authentication, choose No sign in.
Under request headers, add a header named Authorization with the value Bearer YOUR_TOKEN, then click Add.

On Education and Enterprise plans, an organization Owner adds the connector in Organization settings > Connectors first, and members then click Connect. Free plans can add one custom connector. See Anthropic's guide Get started with custom connectors using remote MCP for details.

Cursor

Add the server to ~/.cursor/mcp.json (all projects) or .cursor/mcp.json in a project. Set the environment variable PRINTEROS_MCP_TOKEN to your token, or paste the token in place of the variable.

{"mcpServers":{"3dprinteros":{"url":"https://cloud.3dprinteros.com/apimcp","headers":{"Authorization":"Bearer ${env:PRINTEROS_MCP_TOKEN}"}}}}

ChatGPT

ChatGPT cannot connect to the 3DPrinterOS MCP server yet. ChatGPT only connects to MCP servers that use OAuth sign-in or no authentication, and it cannot send a token header. This article will be updated if that changes.

Other MCP clients

Any MCP client that supports remote servers over streamable HTTP with a custom header can connect. Use the server URL https://cloud.3dprinteros.com/apimcp and send the header Authorization: Bearer YOUR_TOKEN.

Examples

These are examples of what you can type and what the assistant does behind the scenes. You do not need to know the tool names; they are shown so you can follow what the AI is doing.

You type	What the AI does
Which of my printers are online and what are they doing?	Calls `get_printers_list` and summarizes each printer's state, temperatures, and progress.
Upload bracket.gcode to my "Prototypes" project and add it to the queue on Printer A, but don't start it.	Calls `upload`, sets the printer type with `get_printer_types` and `file_update` if the file needs it, then calls `print_file` with only_queue set to 1.
Start the job I queued on Printer A.	Finds the job with `get_printer_jobs`, then calls `job_start`.
Pause the print on Printer A. Resume it. Cancel it.	Finds the running job and calls `job_pause`, `job_unpause`, or `job_cancel`.
How long did my last print take and how much filament did it use?	Calls `get_printer_jobs`, then `get_job_info` for the job.
Share Printer A with student@school.edu so they can only add jobs to the queue.	Calls `share_printer` with the "can queue" role.
Who finished prints last month? Give me a table.	Calls `get_finished_jobs_report` with last month's dates.
Create a member account for new.user@school.edu, add them to the "Robotics" workgroup, and give them 20 more balance.	Calls `create_organization_user`, finds the workgroup with `get_workgroups_simple_list`, calls `add_user_to_workgroup`, then `update_users_balance` with +20. Org admins only.

Be specific about starting prints. If you only want a job in the queue, say so. If you ask the AI to "print" a file, it may queue and start it in one step.

Troubleshooting

401 error or the server does not connect. The token is missing, wrong, or expired, or a newer token was generated. Generate a new token on the Account page and update your AI app.
403 or access error on some tools. If the error says the token lacks a required scope (for example org:read or org:write), the token was created without that permission, or the account is not an org admin. Organization management tools need an org admin account and a token with the matching org permission. Enable a new token with the permission selected and update your AI app.
Error "type_id #1001" when printing a gcode file. The file's printer type is not set. Ask the AI to set the printer type for the file and try again.

Use the 3DPrinterOS MCP server with AI assistants