Import Preview
File:
Words:
Characters:
Choose Voice
Free local speech on this device
Preparing voice engine... The first use may take longer while the speech model is downloaded to your browser.
Voice engine ready
Compare
Generating Speech
Speech Ready
Speech uses a local Kokoro model in your browser. No paid speech API and no API key. The model may download from Hugging Face on first use; your text is not sent to a speech service. API: None.
How to Use Text to Speech
- Type, paste or import text.
- Select a supported language.
- Choose a voice.
- Preview voices if desired.
- Select speed.
- Generate speech.
- Play or download audio.
Can I Choose Different Voices?
Yes. After the voice engine is ready, every English voice from the installed Kokoro model is listed. Preview each one, then generate with the voice you selected. Available voices depend on the selected language and the local model.
Can I Import a Document?
Yes. TXT, Markdown, CSV, DOCX, and PDF are read in the browser. Scanned PDFs without selectable text cannot be spoken without OCR, and this tool does not send files to a cloud OCR service.
Is My Text Uploaded?
No. Text, imported documents, and generated audio stay on your device. The browser may download the speech model from a trusted host on first use.
Does It Require an API?
No paid speech API or API key is required.
Read the original guide: Hear the draft before you record it.