Show HN: Use your OpenAI subscription to create audiobooks
The MCP-based PDF-to-audiobook workflow is plausible, but its 'use your subscription' pitch is undercut by an API-key requirement and unresolved security and voice-rights concerns.
PDF TO MP3 is a Show HN project that connects an AI chat client to an MCP server so users can turn PDFs into audiobooks. The setup uses the user's existing Codex or Claude subscription for document processing, while speech generation is billed separately through an OpenAI or Google API key.
The landing page walks through connecting the MCP server, adding a speech key, and prompting the AI to convert an attached PDF.
Key takeaways
- The service is invoked through an MCP server added to Codex or Claude via Streamable HTTP, then authenticated with a PDF TO MP3 account.
- Document processing is handled by the user's existing AI subscription; speech synthesis is paid separately via an OpenAI or Google API key.
- The workflow asks users to attach a PDF in AI chat and prompt: 'Use the PDF TO MP3 MCP to turn this into an audiobook.'
- The page says users can alternatively connect only their API key, and requires JavaScript to save a speech key.
- It markets 'Make your first audiobook $0' but notes speech generation is billed by Google or OpenAI per usage.
What people think
Analysis details
7 comments analyzed. Analyzed .
The sample selects active discussion branches and includes available parent comments. It may omit parts of the full thread. The analysis included 7 of 11 usable stored comments because of the input limit. Selected themes do not measure community opinion or how common a view is.
API key exposure risk when pasting into third-party pages
One commenter compares pasting an OpenAI API key into someone else's web page to sharing credit-card details, framing it as a compromise risk. A reply recommends local, fully offline audiobook generators for people who do not want to paste OpenAI keys into third-party pages.
Local/offline generators and competing audiobook apps
Replies point to fully offline audiobook generators as alternatives to hosted tools, and another commenter promotes Lissin as an app that turns PDFs or links into podcasts, deep dives, or narration with different voices.
Article, PDF, and link conversion into podcast-style audio
Commenters describe workflows that clean web article text, convert it with TTS, and publish it via cloud bucket/RSS for Apple Podcasts; another built Lissin to turn PDFs or links into podcasts, deep dives, or narration. The RSS/cloud step is described as key for listening on a phone.
Last-mile distribution and voice-quality constraints
A reply notes that skipping the cloud bucket/RSS last mile leaves manual MP3 copying to the phone, which is a key friction point. Another commenter says getting the audio to sound right was always harder and built Lissin around that thesis, offering different voices.
Testing and misleading subscription-versus-API-key claim
A commenter expected 'Use the AI you already pay for' to mean subscription authentication via OpenAI/Codex login or device codes, but found the tool asks for an API key instead. They question whether it was tested and call the framing misleading after a quick look.
Negative social perception of AI-generated audio
A reply says receiving AI-generated audio that someone recommends changes how they view the sender negatively, and that they do not want to read LLM outputs or listen to them.