You have a PDF open. You do not have eyes-on-screen time. The usual options are a cloud reader that wants you to upload the file, a paid app with a free tier that runs out mid-chapter, or the robotic system voice built into your OS.
There is a simpler path: open Junco Reader, drop the PDF in, and press play. Narration runs in your browser with on-device Kokoro voices. No account. Your document is not uploaded to Junco servers.
What you get
| Junco Reader | |
|---|---|
| Formats | PDF, TXT, Markdown, pasted text |
| Price | Free |
| Account | None |
| Where audio runs | Your browser (on-device) |
| Documents leave your device? | No |
How to listen to a PDF in about a minute
- Go to read.tryjunco.com.
- Drop a PDF (or TXT / Markdown), or paste text into the box.
- Download the voice model once when prompted. It caches in your browser so revisits skip the wait.
- Pick a voice and speed, then press play.
That is the whole loop. There is no library to sync, no cloud folder, and no subscription gate on the listen itself.
Why “in the browser” matters
Most “free PDF text to speech” tools are upload-first. That is fine for a public whitepaper. It is a non-starter for a draft contract, a patient handout, or notes you would rather keep local.
Junco Reader parses and synthesizes locally. The only optional network fetch after the page loads (besides fonts) is the voice model when you choose to download it.
When you should use the Junco iOS app instead
Junco Reader is for one-off documents you already have open. The Junco iOS app is for the stream of newsletters that lands in your inbox every morning. Connect Gmail or use a Junco inbox address, and you get a short daily podcast without importing each issue by hand.
Same family of on-device voices. Different job.
FAQ
Is Junco Reader really free?
Yes. The browser tool does not require an account or payment.
Does my PDF get uploaded?
No. Documents you open or paste are processed in your browser.
What voices does it use?
Kokoro-82M, running on-device via WebGPU when available, with a WASM fallback.
Can it read scanned image-only PDFs?
It needs extractable text. Pure image scans without a text layer will not narrate well until you OCR them elsewhere first.
Try it
If your pile is newsletters rather than one-off PDFs, download Junco for iPhone or read more on tryjunco.com.