Otoha395 downloadsRead the current note aloud — device voice, or high-quality neural voices via the free Otoha app — with sentence highlighting and smooth follow-scrolling.
Read the current note aloud in Obsidian, highlighting each sentence in place and smooth-scrolling to follow the spoken line.
Works standalone with your device's built-in speech (no setup, no server, works offline and on mobile). For higher-quality neural voices, install the free Otoha app — a menu-bar app that bundles and runs everything locally (no cloud, no accounts).
From the community plugins browser (once listed): Settings → Community plugins → Browse → search "Otoha TTS" → Install → Enable.
Manually: download main.js, manifest.json, and styles.css from the
latest release into
<vault>/.obsidian/plugins/otoha-tts/, then enable the plugin.
http://127.0.0.1:8765 — all local, no cloud. Then pick the Kokoro engine in
the plugin settings. (Advanced: you can run your own Kokoro server on that port instead
of the app.)This plugin only ever talks to localhost — it makes no external/internet requests:
http://127.0.0.1:8765 (only when you choose the
Kokoro engine) to synthesize audio.127.0.0.1:8767 and posts playback state to 127.0.0.1:8766. These bind to localhost
only and are best-effort (skipped if the port is busy or unavailable, e.g. on mobile).No telemetry, no analytics, no remote code.
Plain JS, no build step. Pure logic (sentence parsing, text matching, scroll math) has unit tests:
npm test
MIT — see LICENSE.