SherpaTTS

An Android text-to-speech engine that uses Piper or Coqui voices and reads aloud with no connection.

Instead of ElevenLabs

f-droid.org/packages/org.woheller69.ttsengine

Screenshot from f-droid.org/packages/org.woheller69.ttsengine, taken 2026-10-30, shown to help you recognise the tool. It belongs to its maker. Ask us to remove it.

Your phone already has a voice that reads articles aloud, and SherpaTTS is how you give it a better one. It is an Android text-to-speech engine based on Next-gen Kaldi that uses Piper or Coqui voices, and it downloads the voice you pick from Hugging Face on first launch. After that the F-Droid page says text to speech works entirely offline.

SherpaTTS scores a 5 on lock-in, a 4 on footprint, self-direction, power and delivery and a 3 on longevity.

One developer, listed as woheller69, makes it under GPL-3.0 only, and version 3.4 arrived on 21 July 2026. F-Droid tags it with a Non-Free Network Services anti-feature for the Hugging Face download, and the permission list includes full network access, running at startup and keeping the phone awake. The page says internet is needed only for the download, so we do not know why the permissions are that broad, and we have not tested it.

What stands out

• Android speech engine using Piper or Coqui voices, offline after a one-time download
• GPL-3.0 only, made by one developer, version 3.4 from July 2026
• F-Droid flags the model download and lists broad network permissions

For comparison, we also reviewed ElevenLabs (16/30) as a benchmark.