A REST API for documents and images, a REST API for forms, and an MCP server so an AI assistant can use either. One API key, one team scope, the same rules everywhere.
POST data to a template and get a PDF or image URL back. Sync, async, bulk, or a Dynamic URL.
Provision forms from code, read submissions back, and get a webhook the moment someone submits.
Connect Claude, Cursor, or any MCP client and let it design templates and render documents in your account.
Whichever way you call it, the request goes through the same service layer as the dashboard. Nothing is possible over MCP that the API would refuse, and nothing costs differently.