PageToVid

npx add-mcp 'https://pagetovid.com/mcp'

What you say, what happens

The film is real: PageToVid made it from pagetovid.com. The conversation is an illustration; the tool names are the ones your assistant really calls.

1. 01 You ask in your own words “Make a 30-second video of our pricing page for LinkedIn.” A link and a sentence are enough.
2. 02 It reads your page PageToVid opens the page in a real browser, picks what to show and writes a storyboard from what the page actually says.
3. 03 It films, voices and renders The real site is filmed, an AI voice reads the script, captions are added, and the MP4 link comes back in the chat, typically in about 6 minutes.
4. 04 Then you edit it by talking Ask for changes and the assistant edits the storyboard and renders the film again: fewer words, another shot, another voice, another language.

Every assistant, step by step

Claude and ChatGPT take the connector URL in their settings; the coding assistants take one command or a few lines of config. Each one signs you in to PageToVid in your browser the first time it connects.

Claude

More about PageToVid in Claude →

1. 1In Claude on the web or desktop, open Customize → Connectors.
2. 2Click “+ Add”, then “Add custom connector”. Name it PageToVid and paste the connector URL above.
3. 3Claude opens a PageToVid sign-in. Sign in (or create a free account) and approve access.
4. 4In a chat, open the “+” menu → Connectors, check that PageToVid is switched on, and ask for a video.

Anthropic offers custom connectors on its Free, Pro, Max, Team and Enterprise plans; a Free account can add one. On Team and Enterprise, an owner first adds the connector under Organization settings → Connectors, then members connect it. Once connected on the web or desktop, it is available in the Claude mobile apps too. Menu names are as they appear in Claude's English interface.

Steps checked in October 2026 against the vendor's own help pages: Claude help: custom connectors using remote MCP · Claude help: using connectors

ChatGPT

More about PageToVid in ChatGPT →

1. 1On chatgpt.com, turn on Developer mode. OpenAI's developer guide puts it under Settings → Security and login; some of its help pages show it under Settings → Apps → Advanced settings.
2. 2Create a new app (in OpenAI's current guide, the “+” button on chatgpt.com/plugins). Name it PageToVid, paste the connector URL above as the MCP server URL and choose OAuth.
3. 3Sign in to PageToVid in the window that opens (or create a free account) and approve access.
4. 4In a new chat, choose Developer mode from the “+” menu, select PageToVid and ask for a video. ChatGPT asks you to confirm before it runs a tool that creates something.

OpenAI's developer guide lists developer mode for Plus, Pro, Business, Enterprise and Edu accounts, on ChatGPT for the web only, not the mobile apps. On Business, Enterprise and Edu, a workspace admin enables it first. OpenAI's pages differ on which plans get full write access, so check its current guide if ChatGPT refuses a tool. Menu names are as they appear in ChatGPT's English interface.

Steps checked in October 2026 against the vendor's own help pages: OpenAI: ChatGPT developer mode · OpenAI help: developer mode and MCP apps in ChatGPT

Claude Code

1. 1Run this command in your terminal.
2. 2Inside Claude Code, type /mcp and authenticate PageToVid: a browser opens, you sign in to PageToVid once and approve.
3. 3Ask for a video in plain words.

claude mcp add --transport http pagetovid https://pagetovid.com/mcp

Or install the plugin: the same server plus slash commands (/pagetovid:video,:animate,:edit,:howto,:show). Type these inside Claude Code:

/plugin marketplace add PageToVid/claude-plugins
/plugin install pagetovid@pagetovid

Cursor

1. 1Click the install button: it opens Cursor with the PageToVid server already filled in.
2. 2Confirm in Cursor; it runs the PageToVid sign-in for you.
3. 3Ask Cursor's agent for a video.

VS Code

1. 1Click the install button to add PageToVid to VS Code (Copilot agent mode), or add the lines below to.vscode/mcp.json.
2. 2VS Code signs you in to PageToVid on first use.
3. 3In agent mode, ask for a video.

{
  "servers": {
    "pagetovid": {
      "type": "http",
      "url": "https://pagetovid.com/mcp"
    }
  }
}

Using Claude Code inside VS Code? The PageToVid Studio extension sets it up for you.

Codex

1. 1Add these lines to ~/.codex/config.toml.
2. 2Codex runs the PageToVid sign-in on first use.
3. 3Ask Codex for a video.

[mcp_servers.pagetovid]
url = "https://pagetovid.com/mcp"

Gemini CLI

1. 1Run this command in your terminal.
2. 2Start Gemini CLI and ask for a video; PageToVid asks you to sign in the first time a tool is called.

gemini mcp add --transport http pagetovid https://pagetovid.com/mcp

Other MCP clients

1. 1Add PageToVid as a remote MCP server (Streamable HTTP) with the connector URL above; most clients accept the JSON below.
2. 2Sign in to PageToVid in the window the client opens.
3. 3A client that cannot open a sign-in can send an API key from your account page as an Authorization header instead.

{
  "mcpServers": {
    "pagetovid": {
      "type": "http",
      "url": "https://pagetovid.com/mcp"
    }
  }
}

Building an agent, a script or a CI job? Send an API key instead of signing in, or call the tools from the Claude API. MCP reference →

What you can ask

Your assistant gets 82 PageToVid tools: it can make a film, edit it, translate it, plan the videos a whole site needs and keep a recurring show going.

• “Make a 30-second video of https://yourcompany.com for a LinkedIn post.”
• “Film our pricing page as a vertical video and mention the free plan.”
• “Show me the storyboard, then make scene 2 say it in fewer words.”
• “Render it again with a calmer voice and our brand look.”
• “Make German and French versions of the film.”
• “Add an AI clip of our product on a desk in scene 4.”
• “Plan the videos our website needs, then make the first two.”
• “Every Monday, make a short video of our newest blog post.”

Editing the storyboard is free, and your assistant can quote a new render before it starts. AI images and clips need a paid plan or a film pass.

Frequently asked questions

Does it work on Claude Free?

Yes. Custom connectors work on every Claude plan, and a Free account can add one. On the PageToVid side every plan works too, Free included: you start with 120 credits, enough for 3 films, and 3 free films per website.

Do I need an API key?

No. Claude, ChatGPT, Claude Code, Cursor, VS Code and Codex sign you in with OAuth: a PageToVid window opens, you sign in and approve what the assistant may do. An API key, from your account page, is only for clients that cannot open a sign-in, such as a headless agent or the Claude API.

Which ChatGPT plans can use it?

Custom MCP apps need ChatGPT's developer mode, which OpenAI lists for Plus, Pro, Business, Enterprise and Edu accounts, on the web. On the PageToVid side, every plan works.

Can my assistant edit a video it already made?

Yes. It reads the storyboard scene by scene, changes the words, captions, shots, visuals, voice, music or look, then renders the film again. It can also translate a finished film into another language. Editing the storyboard is free; the assistant can quote a new render before it starts.

What does it cost?

Every PageToVid tool works on the free plan: 120 credits to start, enough for 3 films, and 3 free films per website across free accounts. AI images and clips need a paid plan or a film pass.

How do I disconnect it?

Remove the connector in your assistant, or end the access from PageToVid: Account → Connected apps lists every assistant you approved, with a button to revoke it.

What is MCP?

The Model Context Protocol: an open standard that lets an AI assistant use another app's tools. PageToVid runs an MCP server, so once you add its connector URL, your assistant can inspect a page, write a storyboard, render a film and edit it. Tools that spend credits ask you first.

Connect it, or start from a URL here

Prefer the website? Paste a page below and get the same film, no assistant needed.

Product names belong to their owners; they appear here only to say PageToVid works with them.