Skip to content

VIZTOOLS AI · MCP

VizToolsAI in your AI chat.

Describe your idea, fill in missing details and receive the result in the conversation. The editor is there when you need hands-on changes.

  1. 01

    Discuss the task

    The chat asks only for missing details: source, format, duration or quantity.

  2. 02

    Review the quote

    The server fixes the model, parameters and coins before starting. Your balance is not permission to spend it all.

  3. 03

    Receive the result

    Images or videos return to a compatible chat, with the original available separately.

Connection diagram, not a screenshot of a third-party app

MCP · OAuth

Connect your chat.

The owner first enables a closed test and adds your account to the test group. Then select a client. You do not need provider API keys.

ChatGPTNot yet tested in this client
  1. 01

    Open settings

    The current OpenAI developer guide uses Settings → Security and login → Developer mode. Availability depends on your account and workspace policy.

  2. 02

    Add a connection

    Open Plugins → plus, enter VizToolsAI and the MCP address above. In the Apps interface, use Settings → Apps → Advanced Settings, then Apps → Create.

  3. 03

    Sign in and review access

    Complete VizToolsAI sign-in and consent. Check permissions and limits. Start by reading your account and model catalog without generating.

ClaudeNot yet tested in this client
  1. 01

    Find connectors

    The connection guide uses Customize → Connectors → plus → Add custom connector. Your organization may require administrator access.

  2. 02

    Enter the server

    Name it VizToolsAI and use the address above for Remote MCP server URL. Check the current Claude guide for the controls available to your account.

  3. 03

    Authorize the connection

    Select Connect, sign in to VizToolsAI and review permissions. Tool access and the inline player are tested separately.

Gemini CLI / MCPNot yet tested in this client
  1. 01

    Check client capabilities

    You need remote MCP with Streamable HTTP and compatible OAuth. The regular Gemini website and Gemini CLI are different products.

  2. 02

    Add the server

    Use the address above. A Gemini CLI configuration example appears below; complete the client’s OAuth sign-in. Do not paste tokens into a chat.

  3. 03

    Test read access

    Ask for your account and model catalog. Terminal clients may return a file link; an embedded video player is not promised.

Gemini CLI example: settings.json
{
  "mcpServers": {
    "viztoolsai": {
      "httpUrl": "https://viztoolsai.com/mcp"
    }
  }
}

01 · READ ONLY

A first request with no spend.

Expect account permissions and model information, or a clear setup message. A health response or tool listing alone does not prove that generation is ready.

Use VizToolsAI to show available models and my connection limits. Do not generate anything or start paid tasks.

Keep the result in the conversation.

Inline images, video, seeking and attachment import depend on the client and need separate tests. When cards are unsupported, use the protected result link. Viewing a completed file or refreshing its link does not start a generation.

Access stays in your control.

  1. 01

    Selected files

    The chat receives only permitted files, not your whole library by default. Upload through VizToolsAI when the client cannot transfer an attachment.

  2. 02

    Clear spending

    Discuss each task’s parameters in the chat. Starting requires server authorization; paid access is currently closed.

  3. 03

    Revoke from your account

    Open Account → AI chats in the editor to review permissions, files and limits, or revoke a connection. Revoking does not erase files already shared in an external chat.