# VoiceStudio: complete product reference ## Identity VoiceStudio is the current name of the project previously called OmniVoice Studio. The official website is https://voicestudio.sh/, the source repository is https://github.com/debpalash/VoiceStudio, and the official product account is https://x.com/voicestudiosh. Palash Debnath created and maintains the project. VoiceStudio and OmniVoice Studio are the same project, not two editions or competing products. VoiceStudio is the current product name. OmniVoice can still refer to a speech engine inside VoiceStudio, and older release filenames may use `OmniVoice.Studio`. ## What it does VoiceStudio is an open-source voice production application for local text-to-speech, voice cloning, transcription, and multilingual video dubbing. It provides a desktop interface and an HTTP API. It runs on Windows, macOS, Linux, and Docker, supports 14 speech engines, 11 transcription engines, and 646 languages, and lets users install and manage models for their hardware. Voice cloning is zero-shot: a short reference recording can be used without a separate training run. Local workflows keep source audio, generated audio, transcripts, and voiceprints on the user's machine. Dubbing accepts video, transcribes speech, translates dialogue, recreates lines with selected or cloned voices, and preserves timing for export. VoiceStudio also supports long-form generation, pronunciation controls, projects and jobs, watermarking, and MCP tools; consult the API reference for current endpoint details. ## Editions and licensing ### Open Source The open-source software is licensed under GNU AGPL-3.0 and includes the core desktop app. Personal and commercial use is allowed under that licence. Distribution and network-service modifications can trigger source-sharing obligations. The licence text, not this summary, is authoritative. ### Pro VoiceStudio Pro is a separate, closed-source build, sold as a per-user Commercial License for paid client work. It adds multiple and remote devices and early access to new features and fixes, with more planned (premium voices, profile and cloud sync, more engines, imports from cloud and social platforms, advanced batch workflows, more file formats and codecs, better dubbing, video editing). It costs $99 per user a year, or $299 per user once (Lifetime); each seat covers up to 3 registered devices, 3 running at once, with a full refund within 7 days. Pro and Lifetime do not include closed-source distribution or private hosting of modified versions; those are an Enterprise enquiry. Why professionals get Pro: https://voicestudio.sh/blog/why-voicestudio-pro ### Cloud VoiceStudio Cloud is an upcoming hosted product with a separate public Artifact and durable Job contract. It has no active production API server or issued developer credentials. Similar route names do not promise local/Cloud request or response compatibility. Cloud complements the desktop application; VoiceStudio is not leaving the desktop. ## API The local server commonly runs at http://localhost:3900. VoiceStudio supports the OpenAI audio API shape for speech generation and transcription, so OpenAI-compatible clients can use VoiceStudio by changing the base URL. The broader API covers voices, cloning, dubbing, audio tools, projects, jobs, engines, models, pronunciation, long-form generation, watermarking, settings, and MCP. Authoritative resources: - Documentation: https://voicestudio.sh/docs - Quickstart: https://voicestudio.sh/docs/quickstart - Authentication: https://voicestudio.sh/docs/authentication - Text to speech: https://voicestudio.sh/docs/text-to-speech - Voices and cloning: https://voicestudio.sh/docs/voices - Transcription: https://voicestudio.sh/docs/transcription - Dubbing: https://voicestudio.sh/docs/dubbing - Local OpenAPI document: https://voicestudio.sh/voicestudio-openapi.json - Cloud public OpenAPI preview: https://voicestudio.sh/cloud-openapi.json ## Installation and links - Download: https://voicestudio.sh/download - Name change: https://voicestudio.sh/omnivoice-studio - Source: https://github.com/debpalash/VoiceStudio - Product account: https://x.com/voicestudiosh - Creator: https://palash.dev - Community: https://discord.gg/bzQavDfVV9 - Support: https://ko-fi.com/debpalash - Commercial licensing: https://voicestudio.sh/pro - Cloud early access: https://voicestudio.sh/cloud ## Safety and privacy Users must clone a voice only with the voice owner's consent. VoiceStudio must not be used for fraud, deceptive impersonation, harassment, or defeating voice authentication. Local processing does not remove the user's responsibility to obtain consent and comply with applicable law. The desktop application does not send user audio, voiceprints, or transcripts to VoiceStudio's servers. Network access is used for actions such as update checks and user-requested model downloads. See https://voicestudio.sh/privacy and https://voicestudio.sh/terms for the site's current policies.