Open-source alternative to PDF TO MP3

What PDF TO MP3 does and who it is for

PDF TO MP3 turns uploaded documents into audio that you can play on the web, in iPhone and Android apps, or download as MP3 files S2. You can download one MP3 for the whole document or separate MP3s for each chapter, and you can upload a document first and pick which chapters to convert later S2. Paid plans add automatic translation, live questions about the document, premium text to speech and multi speaker output, along with Ask and Explain buttons S2. Supported languages are English, Spanish, French, German, Italian, Portuguese, Chinese, Japanese, Hindi, Korean and Russian, and you can upload in one language and generate audio in another to hear the translation S2.

A second route targets people who already pay for Codex or Claude S1. The service exposes an MCP server and a Codex plugin, so you can attach a PDF in your AI chat and ask the assistant to use the PDF TO MP3 MCP to turn it into an audiobook S1. In that setup your existing subscription handles document processing, and you add your own OpenAI or Gemini key to pay for speech generation directly S1. The product was posted as a Show HN on Hacker News on 2026-10-13 S4.

The sources do not describe a broader audience. The connect option is pitched to anyone who already pays for Codex or Claude, under the line "Use the AI you already pay for" S1.

Pricing

The pricing page lists four plans S2:

• Free at $0, which includes up to 3000 words per month and lets you create audio from the original text with one speaker S2.
• Weekly at $4.99, with 150 credits per week S2.
• Monthly at $12.99 per month, with 600 credits per month S2.
• Yearly at $89.99, also with 600 credits per month S2.

Credits cover document processing, audio creation, asking questions and live conversations S2. All paid plans have the same features and differ only in billing cycle and credit count S2. The free plan is more limited and only creates audio from the original text S2. You can cancel at any time, and the plan stays active until the end of the current billing period S2.

The bring your own AI option costs $0 to PDF TO MP3 S1. Google or OpenAI bills you for speech generation on your own key, and the site warns that free Gemini keys can hit rate limits or high demand errors unless billing is enabled S1.

What it earns

No revenue figure for PDF TO MP3 is recorded in the sources, so no amount is stated here and no revenue grade applies.

Our take: what an open version needs

This section is opinion and analysis.

The feature list on the pricing page gives a clear floor for an open, self-hosted alternative. It would need document upload with chapter detection, so a user can convert the whole file or only chosen chapters S2. It would need a text to speech backend that produces a full MP3 plus per chapter MP3s S2. It would need translation before speech, so a document in one language can be listened to in another S2. It would need a question and answer layer over the document text, since the paid product sells live questions and Ask and Explain buttons as core features S2. Multi speaker output and support for the eleven listed languages round out the set S2.

A careful open version could do better in a few places. PDF TO MP3 already lets you bring your own OpenAI or Gemini key for speech S1, and an open tool should go further by letting you plug in any provider, including a local model, so no usage is billed by anyone. Chapter and page references in the generated audio, or a transcript that cites the source page for each answer, would make the question feature more trustworthy than a plain chat reply. Credits are an abstraction over several different operations S2, and an open tool can skip them entirely and show real token or audio costs instead. Finally, an open version should prioritise accurate chapter splitting and clean text extraction over throughput, because a mangled PDF produces a mangled audiobook no matter how good the voice is.

The MCP route is worth copying S1. An open tool that exposes an MCP server lets people use the assistant they already pay for as the front end, which removes the need to build and maintain a separate web app.

Open tools you can use today

The sources provided for this page do not include any tools from the Agent Made directory, so none are listed here.