ZONOS2 logo

MODELS

ZONOS2

by Zyphra

modelsource:zyphra.comtext-to-speechvoice-cloningaudioopen-sourcemoeapache-2.0zyphra
Visit WebsiteDocumentationGitHub

Overview

Open-source MoE text-to-speech model for real-time, high-fidelity voice cloning with Apache 2.0 weights and hosted or self-hosted inference paths.

Details

ZONOS2 is Zyphra’s open-source mixture-of-experts text-to-speech model for real-time, high-fidelity voice cloning. Official sources describe Apache 2.0 weights, a Hugging Face model page, a GitHub repository, and both hosted cloud and self-hosted inference paths. Zyphra Cloud is described as a full-stack platform with serverless inference and dedicated inference capacity for running open-weight models without managing infrastructure.

When to Use

Use ZONOS2 when evaluating open-source text-to-speech models for real-time high-fidelity voice cloning. Use it when you need a model with both self-hosted options via the GitHub/Hugging Face materials and hosted deployment through Zyphra Cloud. Use it when Apache 2.0 model weights are important for your evaluation or deployment review.

Getting Started

  1. Read the official ZONOS2 page at https://www.zyphra.com/our-work/zonos2 for Zyphra’s product overview.
  2. Review the Hugging Face model page at https://huggingface.co/Zyphra/ZONOS2 for model documentation and weights.
  3. Inspect the official GitHub repository at https://github.com/Zyphra/ZONOS2 for self-hosted inference materials.
  4. If you prefer managed infrastructure
  5. review Zyphra Cloud at https://www.zyphra.com/cloud for hosted serverless or dedicated inference options.
  6. Review Zyphra’s terms of use before using public demos
  7. playgrounds
  8. APIs
  9. or evaluation-only features.

Key Features

  • •Open-source mixture-of-experts text-to-speech model.
  • •Designed for real-time
  • •high-fidelity voice cloning.
  • •Apache 2.0 weights according to Zyphra’s ZONOS2 sources.
  • •Available through official Hugging Face and GitHub project pages.
  • •Can be evaluated with self-hosted inference paths or Zyphra-hosted cloud infrastructure.

Capabilities

  • •text-to-speech
  • •voice cloning
  • •real-time audio generation
  • •self-hosted inference
  • •hosted inference

Last updated Jul 31, 2026