ngram for other AI assistants

Turn a chat brief into a video you can review.

Use ngram in any AI assistant or agent that supports MCP. Start with a brief or a public page, check the credit estimate, and review the result before you share it.

Connect any AI assistant

  1. Open your assistant’s MCP or connector settings and add a remote server.
  2. Paste the server URL
    https://mcp.ngram.com
  3. Sign in to ngram when your assistant asks. If it supports sign-in, you don’t need an API key.
Try it

“Check my ngram plan and credit balance.”

Prefer a config file?

Example only. The exact format depends on your client.

{
  "mcpServers": {
    "ngram": {
      "url": "https://mcp.ngram.com"
    }
  }
}
My client requires an API key

Use a key if your client accepts bearer tokens and has no sign-in flow. Creating videos with a key requires a paid ngram plan and available credits.

{
  "mcpServers": {
    "ngram": {
      "url": "https://mcp.ngram.com",
      "headers": {
        "Authorization": "Bearer ngs_your_key"
      }
    }
  }
}
Get an API key
Full setup guide

The full MCP workflow

From source to MP4

Use this path in your assistant after you connect.

  1. 01

    Give it a source

    Share a brief or a public URL. Say who the video is for and what it should explain.

  2. 02

    Approve the request

    Review the model, estimated duration and credit cost before you confirm creation.

  3. 03

    Watch the preview

    Play it in chat when supported, or open the project in ngram. Ask for a revision if needed.

  4. 04

    Export an MP4

    Once you’re happy with the preview, ask for an MP4 export and download the finished file.

What you can do

What works in your assistant

The full video workflow

Bring source files into a project, browse creative options, create the video and ask for changes from the same conversation.

Sign in or use an API key

Sign in to link your ngram account when your client supports it. For a client that sends bearer headers, API-key video creation requires a paid plan and credits.

Preview where your client allows

Watch the preview in chat when your client can show it. Otherwise follow the project link and review the video in ngram.

Export when you are ready

Ask for an MP4 export of the current version at a resolution available on your plan, then download the finished file.

Supported clients

Where you can use ngram

Connect any assistant or agent that supports remote HTTP MCP. Available actions and inline playback depend on the client. Claude and ChatGPT have their own guided setup pages.

AI assistants and agents
Remote MCP · sign in to ngram
Clients with bearer-token auth
API key · paid plan
Claude
Guided setup on the Claude page
ChatGPT
Guided setup on the ChatGPT page
How it compares

When MCP is the right trigger, and when it isn't.

ngram MCP
you are here
REST API
raw HTTP
HTTP webhook
event hook
Zapier
automation
Who triggers itAn AI agent in chatYour application codeAn upstream event in another systemAn automated multi-step Zap
Where you review the requestIn the assistant conversationIn your applicationIn the upstream systemIn the Zap configuration
What you configureOne MCP server addressAn HTTP request and credentialAn event and receiving endpointA trigger and action fields
Best forChat-driven agents and copilotsEmbedding ngram inside your productReacting to upstream system eventsNo-code multi-step pipelines

FAQ

Common questions about the MCP server

Use https://mcp.ngram.com for any assistant or agent that supports remote MCP. Claude and ChatGPT use their own addresses, which their integration pages walk you through.

Still curious? Still curious? Chat with us

Other AI assistants

Connect ngram to your assistant

Choose the setup instructions for your client, sign in and try a brief with a clear audience and a source you can review.