Inaudio – Offline dictation and read‑aloud for the desktop

Description and use

Inaudio is a desktop app that offers offline speech‑to‑text and text‑to‑speech, keeping all audio and transcripts local with a searchable history.

Problem

Users want private dictation without sending audio to the cloud, but most tools ship audio to external services.

Outcome

Inaudio provides a fully offline, privacy‑preserving dictation and read‑aloud experience on Windows, macOS, and Linux, with a searchable history and optional low‑latency ASR models.

Workflow

• Press a keyboard shortcut to start dictation
• Speak into the microphone
• Inaudio transcribes locally using sherpa‑onnx
• The transcript appears in the app and a floating window
• The user can edit or copy the text
• The transcript is stored in a searchable history
• The user can paste text or copy to clipboard and have it read aloud via on‑device synthesis

Evidence

I’ve tried a lot of dictation tools, and almost all of them ship your audio off to someone’s cloud

Source

It works the other way too. Paste text or copy something to your clipboard, and it reads it aloud with on-device speech synthesis

Source

Under the hood it’s Electron, React and TypeScript, with sherpa‑onnx doing the speech work

Source

You download a model once, then everything runs on your CPU

Source

One of the options is NVIDIA’s Nemotron streaming ASR model for English, tuned for about 560 ms latency

Source

Is it perfect? No. This is a brand new project with only a handful of stars, so expect rough edges.

Source