Local TTS: Text To Speech

Local TTS is an offline-first voice reader. Paste, type, or import text and hear natural on-device speech with follow-along highlighting.

No cloud voice API. No account. No ads. After the one-time model download, text to speech works offline. Your notes stay on the phone.

WHY PEOPLE INSTALL LOCAL TTS
• Natural AI voices — Kokoro English (28 US/UK voices) + Supertonic multilingual styles
• Offline text to speech after setup — commute, travel, weak signal
• Private: speech is generated on-device, not sent to a voice server
• Word-level follow-along highlighting while it reads
• Notes library with search, tags, and a simple editor
• Speed and pitch (0.5×–2.0×), ±15s seek, background / lock-screen playback
• Import plain text files (UTF-8)
• Free to use — no ads, no login

MADE FOR
Students, writers, language learners, and anyone who wants to listen to their own text without handing it to a cloud TTS service.

HOW IT WORKS
1. Install and finish the voice model download (Wi‑Fi recommended; about 550–600 MB storage)
2. Create a note, paste text, or import a .txt file
3. Pick a voice (and reading language) and tap play
4. Follow along as words highlight; keep listening in the background

VOICES & LANGUAGES
• English neural voices (Kokoro) with US and UK options
• Multilingual reading styles (Supertonic) across 31 speech languages + Mixed/Auto
• Preview voices before you choose

PRIVACY & OFFLINE
Synthesis runs on your device with ONNX. Internet is only needed to download models the first time. Notes are stored locally and are excluded from Android cloud backup.

REQUIREMENTS
• Android 9 or later
• 64-bit device (arm64 / x86_64) — not 32-bit ARM
• About 550–600 MB free storage for models
• 6 GB+ RAM recommended for smooth speech

HONEST LIMITS
Local TTS is a dedicated AI voice reader. It is not Android’s system TTS engine — other apps and TalkBack cannot use it as the device voice. Scan-to-text, PDF import, and audio export are not included yet. Pronunciation and highlight timing can vary by language and punctuation. Multilingual voices need the Supertonic download to finish.

From Umber Lab — Local TTS: listen privately, offline, on your phone.