Skip to content

Releases: EntranceStarfish/VoiceStudioV2

Release list

1.0.0

1.0.0 Latest
Latest

Choose a tag to compare

View all tags
@EntranceStarfish EntranceStarfish released this 30 Oct 21:56
1.0.0
7cb346f
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

🎙️ VoiceStudioV2 v1.0.0

The first stable release. VoiceStudioV2 is a local AI voice studio: change your voice live, generate speech, clone voices and dub video, all on your own machine. No cloud, no API keys, no uploads.

This release adds two headline features on top of upstream VoiceStudio: real-time voice and an OpenAI-compatible API.

✨ Highlights

⚡ Real-time voice

Speak into your microphone and be heard in another voice, live.

  • Pick your input and output devices, choose a voice or a cloned voice, press Start.
  • Works with Discord, OBS, games and calls through a virtual audio cable such as VB-CABLE.
  • Runs fully offline.

🧰 Everything else you get

Inherited from upstream VoiceStudio and included in this release:

  • Text-to-speech with multiple open-source engines
  • Voice cloning and voice design
  • Multilingual dubbing workspace
  • Model Catalogue to install and switch engines
  • Dictation and transcription

⬇️ Downloads

Platform File
🪟 Windows 10 / 11 (x64) VoiceStudioV2-1.0.0-windows-x64.exe

Windows only for now. macOS and Linux builds are planned.

The installer may be unsigned. Windows SmartScreen can show "Windows protected your PC": choose More info → Run anyway. Download only from this Releases page.

SHA-256 checksums

sha256:272a81b62c6bc7f3b4bc29863a2b0e4e81c999f2eaab648d61b8a4479e9dd649

🚀 Getting started

  1. Download the Windows installer (.exe) and run it.
  2. Open the app and pick a model in the Model Catalogue (the first launch downloads it).
  3. For live voice, open Real-time voice, choose your devices and press Start.
  4. For the API, enable the local server in Settings → API and point your app at http://localhost:<port>/v1.

Full guide: see the README.

⚠️ Known limitations

  • Windows only. There are no macOS or Linux builds yet.
  • Real-time latency depends on your hardware and model. A GPU is recommended for live use.

Found a bug? Open an issue with your OS, hardware, app version and steps to reproduce.

🗺️ What's next

  • macOS and Linux builds
  • One-click audiobook creation (EPUB / PDF / TXT to M4B)
  • Video URL to dubbed video with subtitles
  • Vocal and background separation for cleaner dubbing
  • Global-hotkey dictation with LLM post-processing
  • Meeting transcription with speaker labels

See the full roadmap.

🙏 Thanks

  • The authors and contributors of VoiceStudio, whose work this release builds on.
  • The authors of the open-source speech models that power the engines. Each model keeps its own license.

⚖️ Responsible use

Use voice cloning and voice changing only with voices you own or have permission to use. Do not use this software for impersonation, fraud or deceptive content.

If VoiceStudioV2 is useful to you, a ⭐ helps other people find it.

Assets 3
Loading