Pocket TTS gives your CPU a high-quality voice
Audio2026-04-22
Source: kyutai.org
Kyutai's Pocket TTS shows that natural speech synthesis no longer needs a GPU at all, while Kokoro TTS keeps shrinking the footprint further.
Text-to-speech has quietly become a solved problem at small sizes. Pocket TTS produces natural, expressive speech fast enough to run on a CPU — no accelerator required — which makes it practical for on-device assistants, accessibility and offline tools.
It's part of a wider trend of tiny, capable audio models; Kokoro TTS even runs in a browser tab via WebGPU. The rest are on the audio page.