Skip to content
loudkit

Voice gallery

Listen

Section titled “Listen”

Press play on any voice. Filter by language or presentation, search by name, and expand a card to hear the original recording, see its licence, or copy the line that loads the voice. All 28 voices are included in both model downloads and load by name. Every card has one play button and starts with loudr-1. Use the English model selector above the gallery to hear the same story with turbo, with seed 7 and matched loudness. Other languages play loudr-1.

Other languages play loudr-1.
28 voices
English sample text

The room was quiet when she opened the letter. Outside, the rain had finally stopped. She read the first line slowly, then looked up and smiled. There was still time to change her plans, but for once, she decided to stay exactly where she was.

The ten English voices compare loudr-1 and turbo on the same passage, with seed 7 and matched loudness. Other languages use loudr-1 on their own texts. Voice names are product aliases; presentation labels describe the selected sound, not verified donor gender. Sources, licences and usage.

Voice cloning

Section titled “Voice cloning”

Ten seconds of clean audio is enough. A cloned voice is a file of tensors around 150 KB, not a model.

import loudkit as lk


mine = lk.enroll("my-recording.wav", "loudreader/loudr-1", name="my-voice")
mine.save("voices/my-voice.safetensors")

The weights this needs are published, so the snippet runs today. The method, the audio requirements and the measured similarity are in Cloning a voice. Consent is yours to obtain: Responsible use.