Back to MCP

    MCP Setup Guide

    Step-by-step setup guide for connecting BatchEdits to Claude, Cursor, OpenClaw, and other MCP-compatible AI clients in under a minute.

    Prerequisites

    Note on Authentication: BatchEdits uses API keys or OAuth Client Credentials (Bearer Tokens) for secure remote MCP connections. You can easily generate a key from your settings dashboard.

    Claude Code or Windsurf

    Add this to your ~/.claude/settings.json:

    settings.json
    {
      "mcpServers": {
        "batchedits": {
          "type": "http",
          "url": "https://batchedits.com/mcp",
          "headers": {
            "x-api-key": "your_api_key_here"
          }
        }
      }
    }

    Replace your_api_key_here with the API key generated in settings to access your workspace.

    Claude Desktop

    Add this to your claude_desktop_config.json:

    claude_desktop_config.json
    {
      "mcpServers": {
        "batchedits": {
          "command": "npx",
          "args": [
            "-y",
            "mcp-remote",
            "https://batchedits.com/mcp",
            "--header",
            "x-api-key: your_api_key_here"
          ]
        }
      }
    }

    Requires Node.js installed. Replace your_api_key_here with your custom API key to link Claude to your workspace.

    Custom GPTs & OpenClaw

    For clients that support standard OAuth 2.0 flows (like ChatGPT Custom Actions), use our OAuth endpoints:

    • Authorization URL: https://batchedits.com/api/oauth/authorize
    • Token URL: https://batchedits.com/api/oauth/token

    For terminal clients (like OpenClaw), pass the token securely as an environment variable or config:

    openclaw config set env.BATCHEDITS_API_KEY your_api_key_here
    openclaw config set mcp.servers.batchedits '{"type": "sse", "url": "https://batchedits.com/mcp"}'

    Tool Reference

    Once connected, your AI discovers these tools automatically and knows exactly when to use them.

    upload_videos
    Uploads local binary video files to your secure BatchEdits cloud workspace via streaming curl commands.
    list_actions
    Retrieves all available atomic editing actions (e.g., remove_silence, add_captions, crop) that can be chained together.
    get_styles
    Fetches all your previously saved editing preset styles.
    create_style
    Builds a new reusable style preset by chaining together specific editing actions.
    process_video
    Dispatches an uploaded video and a selected style to the remote background workers for immediate processing.
    list_videos
    Checks the real-time status of your videos (pending, processing, completed, failed) and returns final download URLs.
    download_video
    Downloads the video to your local computer.
    download_videos
    Downloads multiple videos to your local computer.

    Ready to automate your editing?

    Generate your token and give your AI video editing superpowers today.