Printify MCP server for your AI

A Personal Access Token, not OAuth. Printify shows it on one screen — dashboard, then My Profile → Connections — and you paste it into the Personal Access Token field on the connect form. There is no redirect URL to register and no token refresh to manage; the token is sent server-side over HTTPS as an Authorization: Bearer header. After that you pick which shops to connect and which workspace each belongs in, and shop_id is filled in from that choice.

It maps 39 operations across seven modules: shops, the product catalog (blueprints, print providers, variants, shipping), Catalog v2's beta shipping lookups, products, image uploads, orders, and webhooks. That splits 22 read to 17 write, and includes narrower records like get_product_gpsr for a product's EU safety information. Twenty-five tools load by default and the rest stay searchable, so the full surface is reachable without filling your client's context.

Yes, if you let it. submit_order creates an order and submit_express_order does the Printify Express version; send_order_to_production approves an on-hold order and hands it to the print provider; cancel_order stops one. Those calls commit real production work at the print provider, so keep the connection read-only if you only want quotes and reports, or turn on approval-required writes and review each one before it goes through.

Printify works in cents. create_product takes each variant's price in cents, so a $24.99 tee is 2499, and get_order returns total_price, total_shipping, and total_tax in cents too, alongside per-line-item cost and shipping_cost. Tell your agent the unit up front, or ask it to divide by 100 when reporting figures back to you — it is the single most common source of a wrong-looking number.

calculate_order_shipping returns a synchronous quote for an order that does not exist yet and persists nothing, so it works on a read-only connection. get_shipping_info gives cost and handling time for every variant of a blueprint from a specific print provider. Catalog v2 adds beta per-tier lookups — get_standard_shipping, get_express_shipping, get_priority_shipping, get_economy_shipping — plus list_shipping_methods for which tiers a blueprint and provider actually offer.

publish_product starts publishing to whichever sales channel the shop is connected to, and the Product publishing started, Product published, and Product publishing failed triggers tell you how it ended. If the shop is a CUSTOM or external channel, Printify expects the outcome reported back: set_publishing_succeeded, set_publishing_failed, and notify_unpublish when the product is removed at the other end. Skip those calls and Printify's publish state stops matching what is actually live.

No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.

Any client that supports a remote Streamable HTTP MCP server — Claude Desktop, Claude Web, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.

Loading 39 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 25-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. You can switch a connection to the full catalog at any time.

A tool is something your AI calls — there are 39 here. A trigger is something PopMCP watches for you: 14 Printify events that can start work on the PopMCP side. Triggers never enter your AI client's context, and none of them ask you to paste a callback URL into Printify.