Placid MCP server for your AI

One Placid API token. Open your project at placid.app/app/projects, then API Tokens in the left menu, and paste the token. If that page is missing, add the API integration to the project first. No developer app or approval is needed. Each token belongs to one project, so to work in a second project add a second connection with that project's token.

get_template_layers lists the dynamic layers of a template and what each takes: text, an image or video link, a colour or a barcode value. generate_image, generate_pdf and generate_video then take plain values by layer name, write them in the form Placid expects, start the file and wait for the link. A layer name that does not exist is refused before any credit is spent. Placid fills templates and does not design them, so only layers marked dynamic in its editor can change.

Placid charges credits from your plan: one for an image, more above a 4000 px canvas, two for each PDF page, and ten for every ten seconds of video. Reading, managing templates and uploading media spend none. Placid also offers a test mode with free watermarked previews, which its API documents no field for. Its API does not report your balance, so check it in Placid.

Placid renders in the background. The generate tools wait for up to 45 seconds, and if the file is not ready they return its id. wait_for_render then picks it up, so nothing is lost and nothing is charged twice. If a render fails, the AI gets Placid's error code, the layer that caused it and what to do about it.

Yes. Picture layers take public links directly. For a file without one, upload_media stores it on Placid for a limited time and returns a link to use; the AI can send it as base64 or from a link. upload_font adds a custom font once you confirm you have the right to use it.

Three triggers, found by comparing lists on a regular check: a template, a template collection or a custom font added to or removed from the project. Placid has no API for registering webhooks, so these are not instant.

In PopMCP, yes: a connection starts read-only, so the AI can browse templates and check files, and generating, uploads, changes and deletes stay blocked until an owner or admin allows write actions. Placid's own tokens have no permission options, and a token can do everything in its project. Webhook links, storage transfers and PDF merges are only available with write actions on.

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, Claude Code, 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.

Connect Placid in PopMCP first, then run claude mcp add --transport http popmcp https://app.popmcp.com/mcp in a terminal. Inside Claude Code, run /mcp and sign in to PopMCP in the browser window it opens, ticking Placid on the consent screen. The Placid tools are then available in that session, and no Placid credential is stored on your machine.

Loading 34 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 33-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. Every tool stays searchable, and on Scale the owner can load the full catalog for a teammate's connection.

A tool is something your AI calls, and there are 34 here. A trigger is something PopMCP watches for you: 3 Placid 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 Placid.