Elffuss Audio

Audio

Describe what you want to hear and it is made here, in your browser: music, sound effects, or a voice reading your text.

Name the genre, the instruments, the mood and the tempo. The model learnt from descriptions in English.

More settings

The model is tuned for 8 steps. More steps take longer and do not always sound better.

The same description with the same seed gives the same result.

Model: MOSS-TTS-Nano, by OpenMOSS (Apache-2.0). 717 MB. It is downloaded the first time and kept in this browser.

Powered by Stability AI. Stable Audio 3 is used under the Stability AI Community License; its text encoder, T5Gemma, also under the Gemma Terms of Use.

Your audio

What you make shows up here and is kept only in this browser.

Storage

The tools only read from the Elffuss model store and cannot delete from it. To free space, empty it from Claw.