Skip to content
Thursday, 5 November 2026
Tenesys AI News
Subscribe
Voice AI· Important· 🧪 Worth Testing

Gradium launches Phonon, a lightweight on-device TTS model

In short: Gradium has released Phonon, a suite of compact (65-75 MB) text-to-speech models designed to run entirely on edge devices. It supports five languages, operates offline, and performs competitively against models significantly larger in size.

Source: Hugging FaceGradiumPhonon 31m, Phonon 40mOriginal article ↗

This summary was generated automatically by AI from Hugging Face's publication. It is our own text, not a copy of the original — facts, figures and quotes belong to the source, linked above and below.

What changed?

  • 1Released Phonon 31m (English, Apache 2.0) and Phonon 40m (5 languages, non-commercial license).
  • 2Introduced ptts inference framework supporting 10+ backends including CPU, GPU, and NPU.
  • 3Full integration with Hugging Face Transformers v5.21 for batch processing.
  • 4Supports over 100 built-in voices and custom voice cloning.
Phonon 31m, Phonon 40m
ParameterBeforeNow
SizeNot specified65 MB and 75 MB
Supported languagesNot specifiedEnglish, French, German, Spanish, Portuguese

Why it matters

This release shifts TTS capabilities from the cloud to the device, enabling real-time voice interactions without network dependency. It is particularly useful for mobile, embedded, and browser-based applications where latency and privacy are priorities.

What it means for AI agents and contact centers

Phonon offers an alternative to cloud-based TTS pipelines. Test its latency and audio quality against your existing cloud solutions. For contact centers, this could offload simple notifications or voice prompts to the client device, reducing server bandwidth and API costs.

🧪 Worth Testing

Excellent candidate for reducing infrastructure costs and latency in applications requiring offline or localized voice synthesis.

Phonon TTS· New

Sources

  • Hugging FaceOfficialPrimary source
    „A Closer Look at Phonon TTS. Launching High-Quality On-Device Speech That Tops the Benchmarks.“
    5 Nov 2026, 02:00
    Original article →
Published by source
5 Nov 2026, 02:00
Found by our system
5 Nov 2026, 12:37
Summary generated
5 Nov 2026, 12:37

This article was written by AI from the original source. Facts, numbers and prices come from the source; missing values are marked “Not specified”. Legal notice, copyright and privacy

Gradium launches Phonon, a lightweight on-device TTS model · TENESYS AI NEWS