← Back to Model Beat
Open Source·11h ago·all news from October 22, 2026

Gradium Open-Sources Phonon, a 31M-Parameter On-Device TTS Model Under Apache 2.0

Gradium has released its Phonon text-to-speech model under an Apache 2.0 license. The 31M-parameter model is designed for on-device use, requiring 64.5 MB of storage and achieving a 7-millisecond latency on laptop processors.

Covered by 1 source

Related stories

Open SourceA hacker broke into John Deere and SharkNinja with an army of AI agentsOct 19 · 6 sourcesOpen SourceOlmoEarth v1.3: State-of-the-art embeddings for the whole planetOct 22Open SourceMOSS-Image Technical ReportOct 19Open SourceFederated Community Detection in Bipartite NetworksOct 20