Your voice. Any text.

Local voice cloning and text-to-speech for Mac

Seagull Studio is a native macOS app for local voice cloning and text-to-speech. Record a voice sample of at least 10 seconds, write your script, and generate speech on your own Mac. Audio and text never leave the machine, and generation keeps working offline after models download. It runs on Apple Silicon Macs with macOS 14 or later, costs $11.99 once, and has no subscription.

Download on the Mac App Store See how it works →

20 free generations · $11.99 lifetime unlock · no subscription

Simple workflow

From sample to speech in three steps.

  1. 01

    Record or import

    Use a voice sample of at least 10 seconds, recorded in Seagull or imported as WAV, MP3, M4A, AAC, AIFF, or CAF.

  2. 02

    Write your text

    Choose a model, enter your script, and let Seagull detect the language or set it yourself.

  3. 03

    Generate and export

    Create natural speech, keep a history of generations, and export the result as WAV or MP3.

Local-first

Create without sending your voice to the cloud.

Seagull uses a local AI backend on your Mac for voice cloning and text-to-speech. After a model is downloaded, everyday generation can work offline.

  • Voice samples and scripts stay on your Mac for local synthesis
  • Download models only when you need them
  • No recurring cloud-synthesis subscription

Model manager

Choose the voice model for the job.

Download, enable, and configure models in Seagull. Model availability and language support differ by provider.

Apple System Voice

Built into macOS for fast text-to-speech.

System voices

Chatterbox Turbo

Fast English voice cloning with expressive tags.

English · ~4 GB

Chatterbox

Voice cloning across 23 languages.

23 languages · ~13 GB

Qwen3-TTS

Multilingual voice cloning with auto language detection.

10 languages · ~4.5 GB

OpenVoice V2

Cross-lingual cloning and English accents.

6 languages · ~2 GB

Supertonic + OpenVoice

Broad multilingual coverage in a compact download.

31 languages · ~0.7 GB

Built for real voice work.

Manage samples, model settings, generation history, and exports in one native macOS studio.

Language-aware text

Automatic language detection, Russian stress preprocessing, and Japanese kana preprocessing help prepare text for speech.

Local MCP tools

Connect compatible local agents to list voices and models, upload a voice sample, or generate audio.

Lifetime access

Try 20 successful generations free. Unlock all Studio features for $11.99 once, with restore purchases and Family Sharing.

Requirements

Made for Apple Silicon.

FAQ

Questions, answered.

Does Seagull Studio send my voice or text to the cloud?

No. Voice cloning and text-to-speech run locally on your Mac. Internet access is used for model downloads and app access features.

Is Seagull Studio a subscription?

No. You get 20 free successful generations, then can buy a lifetime unlock for $11.99. There is no recurring subscription.

Does it work offline?

Yes. After a model is downloaded, everyday generation can work without an internet connection.

Which Macs are supported?

Seagull Studio requires macOS 14 or later and an Apple Silicon Mac. Intel Macs are not supported.

Which languages are available?

Available languages depend on the selected model. Supertonic + OpenVoice currently offers the widest coverage: up to 31 languages.

Can I use someone else's voice?

Only if you have the rights and permissions required to record, clone, and use that voice. Read the Terms of Service for details.

Notes

Why local, and how it compares.

Ready when you are

Make your Mac speak in your voice.

Get Seagull for Mac