Kyutai Labs has open-sourced Pocket TTS, a 100M-parameter text-to-speech (TTS) model that runs in real time on standard CPUs without a GPU. It delivers audio streaming and voice cloning from a few seconds of sample audio entirely on local hardware, offering a Python API/CLI alongside a browser implementation example.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.