AI Video · MCP

AI Video MCP Server give your AI agent video generation

A Model Context Protocol server for AI video. Connect the Vivideo MCP to Claude, Cursor or any MCP client and your agent can generate video across 30+ models as a native tool — from a prompt, in the flow of work.

Add the MCP server Read the MCP docs

claude_desktop_config.json

{
  "mcpServers": {
    "vivideo": {
      "command": "npx",
      "args": ["-y", "@vivideo/mcp"],
      "env": { "VIVIDEO_API_KEY": "vv_live_..." }
    }
  }
}

Tools your agent can callcreate_auto_videocreate_manual_videolist_modelsestimate_creditswait_for_video

The Vivideo MCP server exposes AI video generation to any MCP-compatible agent. Add it to Claude, Cursor or another MCP client and the agent gains video as a native tool — it can generate text-to-video, image-to-video, avatars and voices across 30+ models by calling the server, no custom integration. It's how you give an AI agent the ability to make video.

Trusted by individuals at

Harvard University

Salesforce

KPMG

P&G

Columbia University

How the MCP server works

Add the MCP server

Add one line of config — npx -y @vivideo/mcp — to Claude, Cursor or any MCP client, with the API key from your Vivideo account.

The agent sees video tools

The agent discovers 13 video tools — create videos, list models, avatars and voices, estimate credits, track renders.

Ask in natural language

Tell the agent what video you want; it calls the MCP tools with the right parameters across 30+ models.

Get the video back

The agent receives a hosted video URL it can use, share or pass to the next step in its workflow.

What the MCP server gives your agent

Video generation as a native tool for any MCP agent.

Capability	What it does
Standard MCP server	Works with Claude, Cursor and any Model Context Protocol client — no custom code.
13 tools, 30+ models	From create_auto_video to wait_for_video — every generation type, every model, callable by the agent.
Agentic generation	Your agent generates video in the flow of a task, not in a separate app.
Same power as the API	Everything the AI video API does, exposed as agent tools.
Guardrails built in	Rate-limited, idempotent and bounded — an agent in a loop can't flood the API or double-charge you.

Developer resources

Everything you need to connect an agent — docs, registry, source.

MCP server docs

Setup for Claude, Cursor and any MCP client, plus the full tool reference.

Official MCP Registry listing

Published as io.github.egocen-vivideo/vivideo — discoverable by any agent.

@vivideo/mcp on npm

One npx command runs the server — nothing to install.

Agent skills

Nine open-source skills that teach agents to use video tools safely.

Open source on GitHub

MIT-licensed — audit every line of the server your agent runs.

Get an API key

Create a key in your Vivideo account and drop it in the config.

How the AI video MCP server works, explained

AI agents are getting good at doing real work — but they can only do what their tools allow. The Vivideo MCP server gives them one more capability: video generation. Add it to an MCP client like Claude or Cursor and the agent can make video across 30+ models as a native tool, calling it in the flow of a task instead of sending you to a separate app.

Model Context Protocol is the open standard that makes this possible. It lets agents discover and call external tools with a consistent interface, so the Vivideo MCP server plugs in without custom integration. Once it's connected, 13 tools — from create_auto_video and create_manual_video to list_models, estimate_credits and wait_for_video — show up as native actions the agent can invoke on its own.

The result is agentic video. Tell your agent to "make a 15-second product teaser and drop it in the deck," and it can generate the clip via the MCP tools, pick the right model, and hand back a hosted URL for the next step — video as a first-class action in an autonomous workflow.

It's the same power as the AI video API, shaped for agents rather than code, and secured with your API key on usage-based pricing. The server is open source (MIT) on GitHub, ships on npm as @vivideo/mcp, and is listed on the official MCP Registry — and nine ready-made agent skills teach your assistant to check credits first, avoid duplicate generations and wait for renders safely. If the API is how your software makes video, the MCP server is how your AI agent does.

AI video MCP server — FAQ

What is the AI video MCP server?

It's a Model Context Protocol server that exposes video generation to AI agents. Add it to Claude, Cursor or any MCP client and the agent can generate video across 30+ models as a native tool.

What is MCP?

Model Context Protocol is an open standard that lets AI agents call external tools. The Vivideo MCP server makes video generation one of those tools, so an agent can make video during a task.

Which clients does it work with?

Any MCP-compatible client — Claude Desktop, Claude Code, Cursor and more. The server is listed on the official MCP Registry and runs from npm with a single npx command.

What can the agent generate?

Text-to-video, image-to-video, avatars and voices across 30+ models — the same capabilities as the AI video API, exposed as agent tools.

How do I set it up?

Point your client's MCP config at npx -y @vivideo/mcp with your API key in the env block — the 13 video tools appear to the agent immediately. Per-client setup lives at developers.vivideo.ai.

Is it secure?

Yes — your key stays in the client's env config and is never logged or returned, and every call is rate-limited and idempotent so retries never double-charge. The server is open source, so you can audit every line it runs.

How is it different from the API?

The API is for code; the MCP server is for agents. It exposes the same generation capabilities as native tools an AI agent can call on its own.

Is the MCP server open source?

Yes — MIT-licensed in the vivideo-toolchain repo on GitHub, published on npm as @vivideo/mcp, and listed on the official MCP Registry, with docs at developers.vivideo.ai.

A developer terminal with an MCP config connecting video tools

Video Generation as an Agent Tool

The Vivideo MCP server exposes video generation to any MCP agent — Claude, Cursor and more. Your agent gains text-to-video, image-to-video and avatars as native tools it can call. The answer to "how do I let my AI agent make video."

Connect once; the agent generates video in the flow of a task.

An AI agent calling video-generation tools via MCP

Built on the Open MCP Standard

Model Context Protocol lets agents discover and call tools with a consistent interface, so the server plugs in with no custom integration — generation just appears as tools the agent can invoke.

Standard MCP in, agentic video out.

Agentic video generation across 30+ models via MCP

The API's Power, Shaped for Agents

Everything the AI video API does — 30+ models, every generation type — exposed as agent tools, secured with your key on usage-based pricing.

If the API is for your code, the MCP server is for your agent.

Everything the MCP Server Gives Your Agent

One server that makes video a native tool for any MCP client.

Works with Claude, Cursor and any MCP client, no custom code.

Create, list, estimate and track — every capability an agent needs.

Video as a first-class action in autonomous workflows.

Rate-limited, idempotent, bounded waits — safe in agent loops.

Who Uses the AI Video MCP Server?

Anyone building with AI agents:

Agent builders

Give autonomous agents the ability to make video.

Developers

Add video tools to Claude or Cursor workflows.

Automation teams

Wire video generation into agentic pipelines.

AI-native products

Let your app's agent produce video on its own.

One Free Studio Behind Every Tool

Each utility here is a doorway into the full Vivideo studio — create, edit and localize video in the same place.

Generate video from an agent

Connect the MCP server and ask your agent for a video — it calls the tools across 30+ models and returns a URL.

Set up the MCP server →

See the MCP Server in Action

Watch an AI agent generate video as a tool — then connect the MCP server on Vivideo.

Loved across every platform

Free AI video generator reviews from real creators

Rated 4.8 out of 5 — 1,624 reviews across Trustpilot, Google Play, App Store, Capterra, G2

Trustpilot AI video generator reviews

Go to review

Genuinely impressed

I've tried a bunch of AI video tools and Vivideo is the first that actually nailed what I described. A one-line prompt turned into a polished clip in minutes, and the avatars and voices feel real.

Dave

Verified review

Give your agent video generation

Connect the Vivideo MCP server and your AI agent can make video across 30+ models as a native tool — in the flow of work.

Add the MCP server Read the docs

API CLI Automation