Top stories
|
RSS
Announcements
Chatter
Podcasts
Policy
The week
Models
Connect your AI
Announcements
Chatter
Podcasts
Policy
The week
Models
Connect your AI
Search
Announcements
Chatter
Podcasts
Policy
The week
Models
Connect your AI
Top stories
Top stories
Phonon · Gradium — forck
|
|
Gradium
All models
Announcements
Chatter
Podcasts
Policy
The week
Models
Connect your AI
Phonon
Announcements
Chatter
Podcasts
Policy
The week
Models
Connect your AI
Search
Announcements
Chatter
Podcasts
Policy
The week
Models
Connect your AI
Gradium models
Phonon
So
far
20 Oct
announced
·
2026
6
posts about it
·
First-party
1
outlet reported
·
On its posts
So
far
20 Oct
announced
·
2026
6
posts about it
·
First-party
1
outlet reported
·
On its posts
Gradium
·
All models
Launch
Gradium
· 5 days ago
Open-Sourcing Phonon: A Suite of Frontier On-Device TTS Models
31M and 40M on-device text-to-speech models, Apache 2.0 and CC BY-NC
Open-weight release
Everything else, by date
Gradium releases Phonon, an open-source text-to-speech model that runs on any device
Open-weight release
21 Oct
Smaller, clearer, better: Phonon Multilingual beats Magpie and NeuTTS Nano in French, German, and Spanish
New model
15 Jul
Phonon update: 1.00% WER on Seed-TTS, smaller than every model we beat
Model update
26 May
Evaluating Phonon: how we made the best TTS model for edge devices
Research paper
9 Apr
Gradium Phonon: On-Device TTS for Consumer Apps, NPCs, and Offline Products
Product launch
3 Apr
About the model
Lab
Gradium
Licence
Apache 2.0; CC BY-NC
Price
a fixed cost per device and model type, with unlimited generations
Weights
Open
Released
20 October 2026, per models.dev
About the model
Lab
Gradium
Licence
Apache 2.0; CC BY-NC
Price
a fixed cost per device and model type, with unlimited generations
Weights
Open
Released
20 October 2026, per models.dev
Phonon