Rainfrog MCP Is Live: Create Campaigns From ChatGPT, Claude, Cursor, and Grok Bot
Rainfrog MCP is live. You can now connect ChatGPT, Claude, Cursor, Grok Bot, Claude Code, and any compatible MCP client to your Rainfrog studio with one public URL and OAuth. No API key. Full setup lives on the Rainfrog MCP setup guide.
This post covers what Rainfrog MCP is, what an agent can do with it, who can use it today, and how the MCP connection differs from the official ChatGPT and Claude plugins that are coming next.
Table of Contents
What Is Rainfrog MCP?
MCP stands for Model Context Protocol. It is an open standard that lets AI assistants call tools on remote servers. Rainfrog MCP is our public Streamable HTTP server at https://mcp.rainfrog.ai/mcp. Add that URL as a custom MCP server or connector in your client, sign in with your Rainfrog account, and the agent can work inside your studio.
Access is OAuth 2.1. The client opens a Rainfrog sign-in page, you approve access, and the client keeps its own token. Never paste a password, token, or service key into chat. There is no separate MCP fee: generations use the same plan and credit balance as the web studio.
What It Does
Once connected, an agent can direct creative work in your Rainfrog account instead of generating one-off images with no memory. In practice that means:
- Campaigns — start a multi-shot set with a locked character, product, style, and direction, then revise individual frames.
- Libraries — create and organize reusable characters, products, styles, environments, and brands.
- Studio workflows — create projects, place templates (Fashion Shooting, Product Still Life, UGC Content), connect nodes, and run image or video generation.
- Interactive studio — in compatible MCP app hosts, open the Rainfrog canvas, inspect the graph, and keep editing visually.
- Export — pull short-lived download links for finished images and videos while source media stays private.
Example prompts that work well: "Create a fashion editorial from these photos," "Make a studio still life of this product," "Turn this look into a short campaign video," or "Regenerate shot 2 with warmer light and share the download link."
Who Can Use It
Anyone with a Rainfrog account can connect. Generative actions need an active plan and consume credits at the same rates as the studio. The same endpoint works across clients:
- ChatGPT — add a custom MCP server from Plugins, choose OAuth, then @ Rainfrog in a new chat. Business and Enterprise/Edu get the full tool set when an admin allows custom MCP; Pro can connect with read and fetch permissions.
- Claude — add a custom connector under Settings → Connectors. Works on Free, Pro, Max, Team, and Enterprise, on web, desktop, and mobile.
- Cursor — one-click install or an mcp.json entry pointing at the Rainfrog URL.
- Grok Bot — ask the Bot to add the custom MCP URL (Grok Bot, not the Grok chatbot), then approve OAuth.
- Claude Code and other MCP clients — any client that supports remote Streamable HTTP with OAuth can use the same URL.
You can revoke a client anytime from Rainfrog Dashboard → Settings → Connected apps, and control which agent action families are allowed from agent permissions.
MCP vs Plugin
Two related surfaces, different jobs:
Rainfrog MCP (available now) is the protocol endpoint. You add https://mcp.rainfrog.ai/mcp as a custom server or connector. One URL works in every compatible client. This is the recommended way to create identity-consistent image and video campaigns from your AI assistant today.
Official plugins (coming soon) are host-specific packages for ChatGPT and Claude directories. The Claude plugin will bundle the same MCP connection with five skills for workspace, canvas, build workflow, art direction, and generation. The ChatGPT plugin will land in the ChatGPT directory. Until those listings ship, use the custom MCP setup — same tools, same Rainfrog account.
Think of MCP as the engine and the plugin as a packaged install experience for a specific host. Connecting MCP now does not block the plugin later; both talk to the same Rainfrog studio.
How to Get Started
Client-by-client steps, install snippets, and FAQs are on the Rainfrog MCP setup guide. The short version:
- Open your AI client’s MCP / connectors / plugins settings.
- Add a remote server named Rainfrog with URL https://mcp.rainfrog.ai/mcp.
- Complete Rainfrog OAuth when prompted.
- Ask the agent to list your Rainfrog projects, then start a campaign or open a studio workflow.
If you build campaigns for brands, this is the path from chat to a consistent, editable Rainfrog project without leaving the agent you already use. Start here: www.rainfrog.ai/mcp.
