AudioDN + AI coding assistants

Tell your AI assistant what audio should do.

You don't need to become an expert in audio processing or media delivery to add audio to your application.

Connect AudioDN's MCP server, describe your feature, and let a compatible AI coding assistant work from our documented API and integration patterns.

This page is about building with an AI coding assistant. Transcription is a separate optional output preset: it turns spoken audio into searchable text and timed captions, and it is what AI Credits pay for. Connecting an assistant here does not spend those credits. Transcription AI Credits

Get connected in three steps.

1. Connect the MCP server.

Recommended: use the hosted endpoint. Compatible clients sign you into your AudioDN account with OAuth — no API key in the config.

Alternative: run local stdio with an API Access key in ADN_API_KEY. Knowledge tools work there without a key.

Hosted endpoint

https://mcp.audiodelivery.net/mcp

Local alternative

npx -y @audiodn/mcp

Connect hosted MCP in Cursor

Opens Cursor’s MCP install flow for the hosted URL. Your client then runs OAuth — this does not write your integration.

Client configuration examples

Cursor (hosted OAuth — recommended)

{
  "mcpServers": {
    "audiodn": {
      "url": "https://mcp.audiodelivery.net/mcp"
    }
  }
}

Other remote clients (Streamable HTTP)

{
  "mcpServers": {
    "audiodn": {
      "url": "https://mcp.audiodelivery.net/mcp"
    }
  }
}

Cursor (local stdio)

{
  "mcpServers": {
    "audiodn": {
      "command": "npx",
      "args": ["-y", "@audiodn/mcp"],
      "env": { "ADN_API_KEY": "adn_..." }
    }
  }
}

Claude Desktop (local stdio)

{
  "mcpServers": {
    "audiodn": {
      "command": "npx",
      "args": ["-y", "@audiodn/mcp"],
      "env": { "ADN_API_KEY": "adn_..." }
    }
  }
}

VS Code Copilot (.vscode/mcp.json)

{
  "servers": {
    "audiodn": {
      "command": "npx",
      "args": ["-y", "@audiodn/mcp"],
      "env": { "ADN_API_KEY": "adn_..." }
    }
  }
}

Claude Code

claude mcp add audiodn --env ADN_API_KEY=adn_... -- npx -y @audiodn/mcp

Codex CLI

[mcp_servers.audiodn]
command = "npx"
args = ["-y", "@audiodn/mcp"]
env = { ADN_API_KEY = "adn_..." }

2. Sign in and approve access.

Hosted OAuth: when your client connects, sign in at account.audiodeliverynetwork.com, pick an organization, and Allow.

That mints a revocable API Access key labeled MCP · … under Settings → API Keys. Delete that key to disconnect.

Local stdio: create an API Access key in the dashboard and set ADN_API_KEY — never paste secrets into a prompt or commit them.

Client-Side Upload and Player keys are scoped app credentials, not the MCP API Access key.

Create an account

3. Describe what you want to build.

Use a starter prompt, then point your assistant at AudioDN’s MCP knowledge tools and your existing stack.

Starter prompt

Help me add audio uploading and playback to this application using AudioDN. First inspect the existing stack and use the AudioDN MCP knowledge tools to verify the correct implementation. Keep the changes minimal, use the appropriate upload and playback authorization flows, protect API credentials, and verify that the feature works.

More ways to help your AI assistant.

Recommended

MCP server

Give compatible AI tools access to AudioDN documentation and supported live API operations.

Optional

Agent Kit

Install AudioDN-specific project guidance and integration validation tools.

npx @audiodn/agent-kit init

Coming soon

AI Plugin

Packaged workflows and guidance for supported AI coding environments.

The same AudioDN. Less integration friction.

MCP assists AI development, but the resulting application still uses AudioDN APIs, components, and standard secure delivery mechanisms.