{"version":1,"type":"story","url":"https://digestai.news/story/gradium-open-sources-phonon-31m-on-device-tts-model-under-apache-2-0","json":"https://digestai.news/story/gradium-open-sources-phonon-31m-on-device-tts-model-under-apache-2-0.json","markdown":"https://digestai.news/story/gradium-open-sources-phonon-31m-on-device-tts-model-under-apache-2-0.md","slug":"gradium-open-sources-phonon-31m-on-device-tts-model-under-apache-2-0","headline":"Gradium open-sources Phonon 31M on-device TTS model under Apache 2.0","summary":"Gradium, a Paris voice AI company spun out of Kyutai, has open‑sourced its Phonon family of on‑device text‑to‑speech models. The English‑only Phonon 31m model contains 31 M parameters, is packaged in a 64.5 MB q8 GGUF file and is released under the Apache 2.0 license. A larger 40 M‑parameter version (74.6 MB) and a multilingual 40 M model covering five languages are offered under a CC BY‑NC 4.0 license. Gradium reports that Phonon 31m achieves 1.70 % word‑error rate on a 500‑sentence hard set, compared with 3.50 % for the Kokoro baseline, and 0.51 % LLM‑WER on a easier Seed‑TTS set.\n\nIt runs on CPUs, phones, Raspberry Pi 5, browsers via WASM/WebGPU and NVIDIA GPUs, with no GPU required for inference. The runtime is available under MIT or Apache‑2.0 and supports Rust, Python, Swift, C/C++ and browser bindings, plus an OpenAI‑compatible server and WebSocket interface.","keyPoints":["Phonon 31m has 31M parameters, 64.5 MB file, 1.70% WER on hard set vs 3.50% for Kokoro","Runs on CPUs, phones, Raspberry Pi 5, browsers (WASM, WebGPU) and NVIDIA GPUs; Apache 2.0 license"],"whyItMatters":"On‑device TTS removes cloud latency and per‑character costs, enabling fast, private speech generation on edge devices and expanding privacy‑preserving AI applications.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["Gradium","Kyutai","NVIDIA"],"models":["Phonon 31m","Phonon 40m"],"people":[]},"firstPublishedAt":"2026-10-22T16:56:12Z","updatedAt":"2026-10-22T16:56:12Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"MarkTechPost","title":"Gradium Open-Sources Phonon, a 31M-Parameter On-Device TTS Model Under Apache 2.0","url":"https://marktechpost.com/2026/10/22/gradium-phonon-31m-on-device-tts","publishedAt":"2026-10-22T16:56:12Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Gradium open-sources Phonon 31M on-device TTS model under Apache 2.0\", 22 October 2026, https://digestai.news/story/gradium-open-sources-phonon-31m-on-device-tts-model-under-apache-2-0","publisher":"Digest AI","title":"Gradium open-sources Phonon 31M on-device TTS model under Apache 2.0","datePublished":"2026-10-22T16:56:12Z","url":"https://digestai.news/story/gradium-open-sources-phonon-31m-on-device-tts-model-under-apache-2-0"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}