Logo
Logo

Gradium develops audio language models designed to deliver natural, expressive, ultra-low latency voice interactions at scale and capable of performing any voice task.

Product

  • Models
  • Pricing
  • API Docs

About us

  • Who we are
  • Blog
  • Careers
  • Contact
  • Press

Socials

  • X
  • GitHub
  • LinkedIn
  • Discord
Background Logo

© 2026 Gradium. All rights reserved.

Terms of ServicePrivacy Policy

GradiumPhononOn-Device Text-To-Speech

Read more →See the benchmarks →

Deploy now.
On all your devices.

Listen now

WeightsWord Error Rate ↓Speaker Similarity ↑
Phonon 20260318~100M1.48%56.37%
Kani TTS 2450M4.97%40.73%
NeuTTS Air552M2.18%47.51%
NeuTTS Nano229M1.71%40.15%
See the benchmarks →
Join the Beta