MCP Connector

5 min readOctober 26, 2026

RenderForm provides an official, hosted Model Context Protocol (MCP) server that lets AI assistants such as Claude, ChatGPT, Claude Code, Cursor, VS Code with GitHub Copilot and Windsurf call the RenderForm API directly from a chat conversation - no code required.

https://api.renderform.io/mcp

The assistant works with the templates you design in RenderForm: it finds a template, fills in its texts, images and colors, and returns a link to the rendered file.

Available tools

ToolWhat it doesAccess
get_templatesList templates, optionally filtered by name or tagRead-only
get_templateGet template details, including the editable properties used in dataRead-only
render_imageRender an image, PDF or video from a templateCreates a file, uses credits
create_screenshotTake a screenshot of a web pageCreates a file, uses credits
create_pdfConvert a web page to PDFCreates a file, uses credits
get_render_resultsList render results, optionally filtered by template or batchRead-only
get_render_resultGet the status and file URL of a single renderRead-only
get_usageGet organization usage: credits, uploads and planRead-only
get_fontsList available fontsRead-only
delete_render_resultDelete a render resultDeletes data
delete_templateDelete a templateDeletes data

Read-only tools can't change anything in your account. Assistants usually ask for your confirmation before they run a tool that deletes data.

Connect your assistant

Add https://api.renderform.io/mcp as a remote MCP server. On first use your assistant opens the browser: sign in to RenderForm, check the organization and approve access. You don't need an API key.

  • Claude (web, desktop and mobile) - open Customize → Connectors, click + Add → Add custom connector, paste the URL and click Continue. Leave the OAuth client settings at their defaults. On Team and Enterprise plans an owner adds the connector in Organization settings → Connectors first, then members click Connect. To use it in a chat, click + → Connectors and turn RenderForm on. Claude documentation
  • ChatGPT - open Plugins, click + → Add custom MCP server, paste the URL and choose OAuth authentication. ChatGPT documentation
  • Claude Code - run claude mcp add --transport http renderform https://api.renderform.io/mcp, then type /mcp in Claude Code to sign in. Claude Code documentation
  • VS Code - run MCP: Add Server from the Command Palette, choose HTTP and paste the URL. VS Code documentation
  • Cursor - add the server to ~/.cursor/mcp.json. Cursor documentation
{
  "mcpServers": {
    "renderform": {
      "url": "https://api.renderform.io/mcp"
    }
  }
}

Example prompts

  • "List my RenderForm templates tagged instagram."
  • "Render my Black Friday banner with the headline '50% off everything' and the photo https://example.com/sneaker.jpg."
  • "Create certificates from the 'Course certificate' template for Anna Smith, John Doe and Maria Garcia, and give me the links."
  • "Render the 'Product teaser' video template as MP4 with the title 'New collection' and tell me when it's ready."
  • "Take a full-page screenshot of https://renderform.io."
  • "How many credits do I have left this month?"

How rendering works

  1. The assistant calls get_templates to find a template and get_template to read its editable properties (for example title.text or photo.src).
  2. It calls render_image with data built from those properties, in the same format as the Render image API.
  3. Images and PDFs come back right away with a link to the file. Videos are rendered in the background: the assistant checks get_render_result until the status is done and then shares the link. See Video templates.

Permissions, credits and data

  • Access is granted per organization. The assistant sees the same templates, render results and usage as a member of that organization, and nothing outside it.
  • Renders, screenshots and PDFs created by the assistant use your organization's credits, exactly like API requests. Repeating an identical render returns the cached file at no cost.
  • Every assistant you approve is listed in Account > API Keys > Connected MCP apps, where you can disconnect it at any time. Access also ends automatically when you leave the organization.
  • RenderForm stores only what the tools need to work: your templates, render results and request logs, the same as for API requests. Your chat conversation isn't sent to RenderForm, only the tool arguments. See our Privacy Policy.

Troubleshooting

  • "OAuth configuration required" or "does not support Dynamic Client Registration" - make sure the URL is exactly https://api.renderform.io/mcp, then restart or reconnect the server. If your client can't sign in with OAuth, use an API key.
  • The assistant uses the wrong organization - disconnect the app in Connected MCP apps, reconnect and pick the right organization on the approval screen.
  • Credit or plan limit errors - check your usage with get_usage or on the Pricing page.
  • Video render fails - videos require a template created in the Advanced Editor. See Video templates.

Using an API key instead

Clients without OAuth support can send your API key (from Account > API Keys) in the X-Api-Key header:

{
  "mcpServers": {
    "renderform": {
      "url": "https://api.renderform.io/mcp",
      "headers": {
        "X-Api-Key": "RENDERFORM_API_KEY"
      }
    }
  }
}

Running the server locally

You can also run the open-source server on your machine with npx. It requires Node.js and an API key:

{
  "mcpServers": {
    "renderform": {
      "command": "npx",
      "args": ["-y", "renderform-mcp", "--api-key=RENDERFORM_API_KEY"]
    }
  }
}

The source code is available on GitHub: github.com/RenderForm/renderform-mcp

Support

Questions or problems with the MCP server? Write to contact@renderform.io and include the assistant you use and the template ID. Service status is available at renderform.statuspage.io.