Give your AI assistant a voice.
Works with any model: Claude, ChatGPT, Gemini, Cursor, Copilot, or a local model in any app that speaks MCP. Then just ask. Your assistant makes the audio, transcribes recordings and pulls YouTube transcripts, using your Audexum credits.
https://audexum.com/mcp
Then just ask.
Plain words. Your assistant picks the tool.
Read aloud
“Read this article aloud and give me the MP3.”
How it connects
- 01
Paste it into any assistant: Claude, ChatGPT, Gemini, Cursor, Copilot, a local model…
- 02
Your assistant gives you a link. Open it, sign in, press Copy token.
- 03
Paste the token back. It connects itself and remembers to use Audexum for every voice job.
Add it to your assistant.
The first time your assistant uses it, a browser window opens: sign in (or create a free account) and press Approve. That's the whole setup.
https://audexum.com/mcpClaude
Open Settings → Connectors → Add custom connector, paste the URL, then press Connect. Works in claude.ai and the Claude desktop app.
Good to know
- What does it cost?
- The same credits as the website and API: 2,000 free credits when you sign up plus 250 every day. V4, the most natural voice, comes with the Creator plan and above. Your assistant tells you when you've used half and most of your credits. See plans.
- What can the assistant do with my account?
- Generate speech, transcribe audio and fetch YouTube transcripts. It cannot see your password or payment details, change your plan, or use voice cloning, batch or audiobooks.
- Where does the audio go?
- Each clip gets a private, unguessable link that works for 30 days, then it is deleted.
- How do I disconnect?
- Remove it from your assistant, or revoke the “MCP: …” key under API keys.
- I'd rather use an API key.
- Any client that lets you set headers can send
Authorization: Bearer sk_live_…with a key from the developer page. No browser step.
Claude, ChatGPT, Gemini, Cursor and Copilot belong to their owners. Audexum is independent and not endorsed by or affiliated with any of them.