🗣️ Indic Text-to-Speech

Generate speech with Google gTTS or a larger Microsoft Edge Tamil voice catalog, then optionally clone its tone color from a consented reference recording.

Select one provider checkbox. Edge TTS offers eight Tamil neural voices.
Voice / Variant uses named voices or accent/domain variants where available.
Available languages depend on gTTS support.
Long text is split internally and merged automatically for reliability.
Voice cloning accepts generated or uploaded source audio up to 30 seconds long.
Very large text may take longer to process.

Source speech

Source Mode

Text to Speech Upload / Record Audio

Language

Google gTTS

Microsoft Edge TTS

Voice / Variant

Text

Generated Audio

Target / reference voice

Only use voices you own or have permission to convert. Disclose synthetic output.

Target voice / preview

I confirm I have permission to use the source and reference voice for this synthetic conversion.

Voice converter

UAT comparison: no quality winner selected. Use 0.5–30 seconds of source and 0.5–20 seconds of reference audio. MeanVC2 uses ZeroGPU; ONNX engines use CPU and download on first use.

Voice Converter

Cloning Password

Converted audio

Conversion info