Gradium open-sources Phonon, a text-to-speech model suite under 50M parameters

Gradium releases Phonon, a suite of open-source text-to-speech models with under 50 million parameters each, and says the suite sets a new standard for naturalness and accuracy across languages with the broadest inference support to date.

Phonon-31m is available under the Apache 2.0 license. Modal says it built a voice agent with the new TTS models and Pipecat on its platform, and a demo is available at gradium.ai/phonon.

The report is based on posts from Gradium's and Modal's accounts on X. Gradium's naturalness and accuracy claims are its own and have not been independently measured in the reports.