vocalove

Zonos2 — Multilingual Voice Cloning

Zero-shot clone from a short clip — Zyphra’s open-source Zonos2 speaks naturally across many languages. Apache-2.0 model on fal — pick Clone voice in the tool below. No API key needed here.

Voice
Script
Voice model
0 / 1000

Free to preview in your browser — no credit card. New accounts start with 5 free credits that never expire, plus a free daily pool. Watermark-free HD exports use credits only when you're ready to download.

What is Zonos2?

Zonos2 is Zyphra’s open-source text-to-speech model for high-fidelity zero-shot voice cloning — a short reference clip plus a script is enough to speak new lines in that voice.

On Vocalove, clone mode runs Zonos2 in the cloud — upload or record a clear sample, type what should be said, and generate downloadable speech in the browser. No local GPU or API key setup.

Supported text-normalization languages include English (US/UK), French, German, Spanish, Italian, Portuguese (Brazil), Japanese, Mandarin (cmn), and Korean.

  • Zero-shot clone

    About five to thirty seconds of clear speech sets the voice — no per-person fine-tune.

  • Multilingual scripts

    Write in English, Mandarin, Japanese, Korean, and other supported locales for natural delivery.

  • Minute-based credits

    About 3 credits per minute of generated audio — short lines stay cheap.

  • Open source

    Zyphra Zonos2 (Apache 2.0) — natural speech with MoE architecture, hosted via fal.

Try Zonos2 on Vocalove in three steps

Clone mode is selected for you — sample, script, generate.

  1. Step 1

    Record or upload a sample

    Use 5–30 seconds of clear single-speaker speech (mp3, wav, or m4a). Quiet background works best.

  2. Step 2

    Type your script

    Up to 1,000 characters per generation. Match the language of the line you want spoken — English, Mandarin, Japanese, Korean, and more.

  3. Step 3

    Generate and download

    Preview in the browser, download MP3, or continue to Talking Photo Videos when you want a talking portrait.

Jump to the speech tool ↑

What Zonos2 is good for

When you need a specific voice across languages without training a custom model.

  • Multilingual narration

    Clone once, then write scripts in supported languages for localized lines.

  • Memorial & tribute audio

    Short personal samples become new spoken messages — only with clear permission.

  • Talking photo prep

    Generate speech here, then animate a portrait on Talking Photo Videos.

  • Demos & product voiceovers

    Fast iteration for demos, agents, and short-form clips without a studio session.

Zonos2 on Vocalove

What the browser tool accepts and returns.

Model
Zyphra Zonos2
Text per generation
Up to 1,000 characters
Languages
en_us, en_gb, fr_fr, de, es, it, pt_br, ja, cmn, ko
Clone sample
5–30 seconds of clear speech
Credits
~3 credits per minute of output audio
Output
WAV/MP3 audio for preview and download

Zonos2 FAQ

What is Zonos2?

Zonos2 is Zyphra’s open-source multilingual TTS with zero-shot voice cloning. On Vocalove you use it as Zonos2 — one short reference clip plus a script produces new speech in that voice.

How much does Zonos2 cost in credits?

About 3 credits per minute of generated audio (fal list price $0.01/min at 3× cost). Very short lines often land at 1 credit. See Pricing for packs.

How long should my clone sample be?

Five to thirty seconds of clear single-speaker speech works best. Avoid music, heavy noise, and overlapping voices.

Which languages does Zonos2 support?

Text normalization supports en_us, en_gb, fr_fr, de, es, it, pt_br, ja, cmn (Mandarin), and ko. Vocalove auto-picks Mandarin, Japanese, Korean, or English US from your script when possible.

Can I try Zonos2 for free?

You can preview in the browser. New accounts start with 5 credits that never expire, plus a free daily pool; watermark-free downloads use credits when you download.

Does Zonos2 make talking photo videos?

No — it outputs speech audio only. Animate a portrait on Vocalove with Kling Avatar after you have an audio track you like.

What to do after speech audio

Download when the line sounds right. For a talking photo video, open Talking Photo Videos.

Make a talking photo videoBack to the speech tool