Gradium Open-Sources Phonon, a 31M-Parameter On-Device TTS Model Under Apache 2.0
Gradium has open-sourced Phonon, a family of small text-to-speech models built to run on the device.
Proof1 independent outlet1 community thread
Key points
- The Gradium Phonon on-device TTS model comes in 31M and 40M sizes.
- Performance: top English Elo among open models Gradium tested; about 7 ms to first audio on CPU.
- Phonon is an open-weight TTS family from Gradium, a Paris voice AI company spun out of Kyutai.
- Phonon 40m: 1 model per language (English, French, German, Spanish, Portuguese), CC BY-NC 4.0.
Sources (2)
- [1]Gradium Open-Sources Phonon, a 31M-Parameter On-Device TTS Model Under Apache 2.0MarkTechPost · Oct 22, 04:56 PM
Gradium has open-sourced Phonon, a family of small text-to-speech models built to run on the device.
The Gradium Phonon on-device TTS model comes in 31M and 40M sizes.
- [2]Phonon is now open-source, 31M and 40M TTS models designed to run on the edger/LocalLLaMA (top, daily) · Oct 20, 04:01 PM
Extractive summary: sentences quoted from the sources.
Before this
- Oct 20, 2026Voxtral 3
- Oct 20, 2026[AINews] DeepSeek DeepInfer: Open Source SOTA Inference running 155T toks/day/cluster making $22m/day at ~80% margin for a 81 day payback
- Oct 20, 2026Do Diffusion LLMs Scale? Testing Mercury 3 + Decide
- Oct 20, 2026MPS in PyTorch 2.15
- Oct 19, 2026NVIDIA Cosmos3-H-Dreams: Bringing Clinical Surgery to Life Through Generative Simulation
- Oct 19, 2026NVIDIA Open Sources Run:ai Under Apache 2.0 for Kubernetes GPU Orchestration
