Skip to connection setup
v4v.ai homeConnect your assistant
V4V MCP PREVIEW

Choose your AI assistant.

Your conversation. Your creative studio.
Connect V4V to create video, images and audio with the assistant you already use.

SERVER URLhttps://v4v.ai/api/mcp

Sign in with your V4V account, or use an API key.
Same V4V credits. Same V4V library.

VS Code

Use V4V tools with GitHub Copilot in VS Code.

AUTHENTICATION
  1. 1

    Click Install in VS Code and let your browser open the desktop editor.

  2. 2

    Review and trust the V4V server when VS Code asks.

  3. 3

    Start the V4V server in MCP settings, then sign in and accept the V4V permissions.

  4. 4

    Ask your assistant: “Check my V4V balance.”

.vscode/mcp.json
{
  "servers": {
    "v4v": {
      "type": "http",
      "url": "https://v4v.ai/api/mcp"
    }
  }
}

Requires the desktop editor. If the link is blocked, copy the configuration into .vscode/mcp.json.

V4V SIGN-IN · OAUTH

One sign-in.
All your creative tools.

Connect your existing V4V account. Review the assistant and approve access once—there is no key to create, copy or store.

  • Connecting does not spend credits.
  • You approve access to viewing, editing and generation.
  • Disconnect any time in your V4V profile.
Manage your connections
EXAMPLE
V4V · Assistant connection

Connect your assistant to V4V

Signed in to your V4V account.

Required permissions

View your balance, models, projects and media.Required

Upload media, preview costs, edit and share projects.Required

Generate media and transcribe using your credits.Required

Connecting does not spend credits. Generations use your existing V4V balance. Access expires after 30 days.

Accept all permissionsCancel
API KEY · STEP BY STEP

A key, if that works better for you.

Create it once in V4V. Use it in a client that supports custom request headers.
Here is the complete setup, with Claude as an example.

01IN V4V

Sign in to your V4V account

Use Google or Telegram. Your assistant will use this same account, balance and media library.

Sign in to V4V
View example
V4V sign-in screen with Google and Telegram options
V4V sign-in screen with Google and Telegram options
02IN V4V

Create an API key

Open your profile, find API Keys, give the key a name such as Claude, then choose Create API key.

Open your API keys
View example
The API Keys section and Create API key button in a V4V profile
The API Keys section and Create API key button in a V4V profile
03IN V4V

Save both secrets

Copy the API secret that starts with v4v_live_. Keep it and the webhook signing secret somewhere safe; they are shown only once. Your assistant needs the API secret.

Keep the API secret private. Paste it only into your client’s authentication settings.

View example
Save your API secret and webhook signing secret; example values are hidden
Save your API secret and webhook signing secret; example values are hidden
04IN YOUR ASSISTANT

Add a custom connector in Claude

Open the custom connector form with the button below. V4V and the server URL are filled in for you. Other clients have their own MCP settings.

Add to Claude
View example
The connectors menu in Claude
The connectors menu in Claude
05IN YOUR ASSISTANT

Enter the name and server URL

Name the connector V4V and enter https://v4v.ai/api/mcp. Choose Streamable HTTP when asked for the transport.

https://v4v.ai/api/mcp
View example
The custom connector name and server URL fields in Claude
The custom connector name and server URL fields in Claude
06IN YOUR ASSISTANT

Choose No sign in and add your key

Under Request headers, choose x-api-key and paste the API secret. Header authentication depends on your Claude account or organization. Use OAuth if this option is unavailable.

x-api-key
View example
Claude request header settings for x-api-key authentication
Claude request header settings for x-api-key authentication
07IN YOUR ASSISTANT

Connect and check your balance

Choose Add and enable the tools you want Claude to use. Ask “Check my V4V balance.” Add credits in your V4V profile when you are ready to generate.

View example
Claude showing a connected V4V connector
Claude showing a connected V4V connector

Screenshots are examples. Labels and available authentication options may vary by client version or account.

YOUR FIRST CONVERSATION

Connected. Now make something.

Start with a balance check, then bring your next creative brief.

Start here

Check your connection

Check my V4V balance and list the video models that accept a product image as reference.

See the guide
Video

Turn a product URL into a video

Plan a 10-second vertical 9:16 ad from this product URL. Show me the concept and estimated credits before generating.

See the guide
UGC

Find a new creative angle

Create two UGC-style concepts for this product: a natural talking-to-camera review and a product demonstration. Show costs and wait for my approval.

See the guide
Catalog

Create a product video batch

Prepare one Studio video for each of these product URLs. Show the total credit estimate and ask for my batch budget before starting.

See the guide
Editor

Bring your clips together

Combine my last three V4V clips into a vertical video. Add a title for the first two seconds and show me a preview before export.

See the guide
Audio

Give your ad a voice

Write a 15-second voiceover for this product ad in English and Spanish. Show available voices and the credit estimate before generating.

See the guide
GO A LITTLE DEEPER

Everything else
is in the docs.

Tools, reference media, credit estimates, batch workflows and editing. The full reference, when you need it.

Open MCP documentation Looking for the REST API?
A FEW THINGS TO KNOW

Before you connect.

What is V4V MCP?

V4V MCP connects your AI assistant to your V4V account. It can discover models, use reference media, generate video, images and audio, edit videos and retrieve results through the Model Context Protocol.

Does connecting cost credits?

Connecting, checking your balance, browsing models and checking task status do not spend generation credits. Generations use your existing V4V balance. Ask for an estimate and approve the brief before starting.

Do I need an API key?

Use V4V sign-in (OAuth) when it is offered for your assistant. API keys are an alternative for clients that accept custom request headers. ChatGPT custom MCP connections use OAuth here, not a V4V API key.

Which assistants can connect?

Use the guides for ChatGPT, Claude, Claude Code, Cursor, VS Code, Codex CLI, Gemini CLI, Windsurf, Cline and OpenCode. Other clients can connect if they support remote Streamable HTTP with a custom authentication header.

Where do my results appear?

Your assistant can return result links and task status. Generations also appear in the same V4V account and library used by the website, so you can continue working with them in V4V.

Can I disconnect an assistant?

Yes. Open your V4V profile and disconnect it under Assistant connections, or revoke its API key. This stops new access; generations already submitted continue running.

V4V, WHEREVER YOU WORK

Your assistant.
Your video studio.

Pick your assistant. Connect your account. Bring an idea.

Connect your assistant Read the docs