Back to Model
S

SupraTTS-1.0-Preview

Model Oct 17, 2026

A downloadable English text-to-speech preview with a declared 1.03-million-parameter three-stage core and a local Python runtime.

Visit primary link 2 sources attached
SupraTTS author's three-panel chart plots automatic speech scores against parameter counts, with separate markers for comparisons using different evaluation sets.
Author-published automated-scoring comparison, not an AITrending reproduction or human listening test. Full chart and protocol notes preserved. Some rows use different evaluation sets; the chart's shared-front-end footnote also differs from the current card's bundled-front-end description, so it is not a verified comparison of total installed system sizes. Source: SupraLabs / arthurblg1802
self-hosted

What happened lately

SupraTTS-1.0-Preview timeline

SupraTTS-1.0-Preview converts English text into 24 kHz mono speech through duration, acoustic and waveform-decoding stages. Its released configuration declares 1,031,754 parameters across that core. A bundled English front end handles grapheme-to-phoneme conversion separately.

The author reports fast CPU synthesis and automatic quality scores on a small evaluation set. These are author-run measurements, not an independent listening test. Hub history begins October 15, 2026, with a private-labelled initial release; the exact transition to public availability is not established by those commit timestamps. The October 17 card update does not imply new model weights.