How to use the AI voice generator
- Enter or paste the text you want spoken.
- Choose an American or British English AI voice.
- Select Slow, Normal or Fast speaking speed.
- Click Generate AI voice and wait for the audio preview.
- Listen to the preview, then click Download MP3 to save your audio.
Different AI voices
Choose from American and British English character voices, including female and male options. Voice names identify the available model voices; they are not voice cloning or recordings of a specific person.
Privacy and requirements
AI speech generation runs in your browser using the Kokoro open model. Your entered text is processed locally by the model and is not sent to a Tolvix speech server. Your browser downloads the model and supporting libraries from their public hosts. The first model download is approximately 90 MB. Generation requires a modern browser, an internet connection for initial downloads and sufficient device memory.
See the Privacy Policy and Terms of Use.
Frequently asked questions
Can I download the generated voice?
Yes. Generate the audio first, preview it in the player, then click the separate Download MP3 button. Generation does not automatically download a file.
Can I change speaking speed?
Yes. Choose Slow (0.7×), Normal (1.0×) or Fast (1.5×) before generating audio.
Which voices are available?
The generator includes multiple American and British English female and male voices. Voice names are labels for model voices, not voice cloning.
Why does the first generation take longer?
The browser needs to download the AI model the first time. The model may be cached, so later visits can load faster.