Zyphra Releases ZONOS3-Mini Open-Source TTS Model
Zyphra launches ZONOS3-Mini, an open-source 100M active parameter text-to-speech model featuring voice cloning and real-time browser execution.
Stock photo for illustration only, not from the actual event
- Zyphra has released ZONOS3-Mini, an open-source text-to-speech model under the Apache 2.0 license.
- The model features 800 million total parameters with 100 million active parameters.
- It supports voice cloning, word-by-word delivery control, and real-time execution directly in the browser.
Zyphra has officially introduced ZONOS3-Mini, a cutting-edge text-to-speech artificial intelligence model released under the Apache 2.0 open-source license. This release grants developers and researchers the freedom to utilize, modify, and integrate the technology into various applications.
From a technical standpoint, the model is engineered with a total of 800 million parameters, featuring 100 million active parameters that strike an optimal balance between high-fidelity voice output and computational efficiency.
Stock photo for illustration only, not from the actual event
In terms of practical capabilities, ZONOS3-Mini introduces advanced features including precise voice cloning, granular word-by-word delivery control, and the capability to run seamlessly in real time directly within web browsers.
The decision by Zyphra to license ZONOS3-Mini under Apache 2.0 represents a significant milestone for accessibility in speech synthesis technology. By maintaining only 100 million active parameters while retaining advanced voice cloning and browser-based real-time performance, the model highlights a clear industry shift toward efficient Edge AI solutions that minimize heavy server-side infrastructure demands.
Source: MarkTechPost
Found something wrong in this article? Report an issue with this article
Comments
Leave a Comment