Skip to content

Back to the catalog

Show HN: Use your OpenAI subscription to create audiobooks
pdftomp3.comtool

Show HN: Use your OpenAI subscription to create audiobooks

PDF TO MP3 is a hosted MCP server (Streamable HTTP, account authentication) that turns PDFs into audiobooks. The user's existing AI subscription, such as Codex, handles document processing through the MCP connection. Speech generation is billed separately through the user's own OpenAI or Google text-to-speech API key. The setup is three steps: add the MCP server, add a speech key, then attach a PDF and ask the assistant to use the MCP. It shows a "bring your own subscription" pattern that keeps the vendor's compute costs low.

TL;DR MCP server that converts PDFs to audiobooks, using your Codex/ChatGPT subscription for processing and your own OpenAI or Google key for TTS.
Open original ↗
Document

What it claims

PDF TO MP3 states that it charges $0 when you connect it to an AI subscription you already pay for. The only per-use cost is speech generation, which Google or OpenAI bills directly against your own text-to-speech API key ("OpenAI charges per usage"). Document processing runs on your existing subscription through MCP and the page describes it as "already paid by your subscription."

How to use it

The page gives the setup for Codex specifically:

  1. In Codex, open Settings → Plugins → Add MCP server.
  2. Choose Streamable HTTP and paste the PDF TO MP3 server address. The fetched text does not include the address itself, so get it from the live page.
  3. Save, then choose Authenticate and connect your PDF TO MP3 account. You need an account with the service.
  4. Add your speech key. The page says this key "pays for speech only."
  5. Attach a PDF in the AI chat and send the suggested prompt verbatim:
Use the PDF TO MP3 MCP to turn this into an audiobook.

The page also offers a second path, "Only want to connect your API key? Use my API key", which skips the MCP and subscription. The fetched text does not say how that mode is priced or who does the document processing in it.

Limits

  • Only Codex is documented. The setup instructions name no other assistant, so check that your client supports Streamable HTTP MCP servers with authentication before relying on this.
  • Several things are not stated on the page:
  • which OpenAI or Google voices or models are used
  • any limits on PDF size or page count
  • the output format or length
  • how audio is delivered or stored
  • what the service keeps of your documents or your API key
  • Your API key goes to a third party. You hand your OpenAI or Google speech key to the vendor, and the page gives no detail on how it is stored or scoped. Consider using a restricted key or a key with a usage cap.
  • The "$0" applies only to the vendor's own fee. It does not cover your subscription's usage limits, which a long PDF processed through MCP may consume.

Organisations named here

  • OpenAI
  • Google
  • PDF TO MP3
  • Model Context Protocol

Technologies named here

  • PDF TO MP3
  • Codex
  • MCP
  • Model Context Protocol
  • OpenAI API
  • OAuth

Related in mcp