VoiceStudio

Overview

VoiceStudio is a local studio for voice creation and audio production, suited to creators and developers who need several speech workflows in one place. Its breadth is a strength, but running models locally calls for capable hardware and attention to each model’s licence.

The maker says recordings, generated audio, transcripts, and derived voice data remain on storage users control rather than being received by its website or control plane. The app can still connect to check for updates, download models, or use a remote adapter that the user has configured. Cloud is in early access; the public build does not include the hosted dashboard or Cloud API.

Key features

Voice creation and control

Voice design starts with a written description: users can specify gender, age, accent, pitch, and emotion without supplying reference audio. Voice cloning is supported, with the method described as instant, and pronunciation controls help shape how particular words are spoken. These tools make the studio useful for both designed voices and cloned ones, though the speech models carry separate licences and some may be limited to research use.

Dubbing, transcription, and long-form audio

The video-dubbing workflow transcribes and translates footage, then revoices it while keeping speakers separate and aligning timing with the original. That combination suits projects where preserving speaker turns and timing matters, rather than simple voice-over generation. VoiceStudio also supports dictation and transcription, multi-voice audio from scripts, and chaptered audiobooks made from long text or EPUB files.

Local API and engines

The desktop app exposes an OpenAI-compatible API at http://localhost:3900/v1, with speech, voices, transcription, and dubbing endpoints. The maker lists 26 adapters, spanning local speech and transcription engines as well as a configured-remote OpenAI-compatible speech-recognition adapter. This gives developers a way to connect compatible software to a local installation, but it is not the hosted Cloud API.

Hardware and licensing

The download FAQ says to allow about 8 GB of RAM and around 10 GB of disk for models, recommends 16 GB or more of RAM, and notes that GPU use is optional but CPU processing is slower. VoiceStudio may be used at work or for paid work under AGPL-3.0; speech models have their own terms, so commercial users need to check the licence for each model they choose.

Pricing

The Open Source plan costs nothing under AGPL-3.0, with no seat count or evaluation period; speech model licences are separate. It is the natural starting point for individuals who want the public build and can manage its local hardware and licensing requirements.

Pro costs $99 a year, billed at $99 per user / year. It covers 1 user, 3 active devices, and 3 concurrent uses. Lifetime costs $299, paid once, billed at $299 per user one-time, with the same 1-user, 3-device, and 3-concurrent-use limits. Those paid plans make sense for an individual who wants their stated device and concurrency allowance; the free plan remains the option with no paid plan price.

Enterprise has custom pricing and flexible terms for larger teams. An enquiry is not a purchase, quote, agreement, or licence grant, so it is a conversation rather than a confirmed team licence. There is no free trial.

Platforms

VoiceStudio is listed for Linux, macOS, Windows, web, API, and self-hosted use. Its local desktop app is the basis for the local API; the web platform should not be confused with the hosted dashboard, which the public build does not offer.

Who it's for

VoiceStudio fits creators working across cloning, voice design, dubbing, transcription, and audiobooks, as well as developers who want an OpenAI-compatible local API. It is less suited to users who want a hosted dashboard in the public build, need processing without local hardware demands, or do not want to evaluate model-by-model licensing.

Pros and cons

• Pro: One studio covers voice design, cloning, dubbing, dictation, transcription, and audiobook creation, making it relevant to mixed speech-production workflows.
• Pro: User-controlled storage and a local API support privacy-conscious local use and software integration.
• Pro: Voice design works from text without reference audio, and pronunciation controls add a further way to shape speech.
• Con: The public build lacks the hosted dashboard and Cloud API, so it is not the choice for users seeking the cloud product.
• Con: The recommended 16 GB or more of RAM and slower CPU-only processing may be a hurdle on modest machines.
• Con: AGPL-3.0 does not settle the terms for speech models; users must check each model’s separate licence.

Alternatives

For a narrower choice, Voice Cloning Software, AI Voice Cloning Software, Voice Changer Software, and Text-to-Speech Software are category lists for comparing options by workflow.

• Fish Audio has a free tier with 8,000 monthly credits, up to 7 minutes of generation, up to 500 characters per generation, and 3 public voice slots.
• Voicebox is a freemium alternative with a free plan and no free trial, and supports Android, iOS, API, Linux, macOS, self-hosted, and Windows platforms.
• Fish Speech is free and provides 8,000 monthly credits, up to 7 minutes per generation, up to 500 characters per generation, 3 public voice slots, and standard generation speed.
• Kits AI offers a free monthly plan with 15 conversion minutes, no voice slots or download minutes, plus voice designer/blender and generative vocals.
• Qwen3-TTS offers Apache-2.0 open-source models in 0.6B and 1.7B sizes for 10 languages.
• Sprag is a paid alternative with Chorus stock-voice text-to-speech priced at $0.70 a month, billed at $0.70 per hour of speech generated.
• Chatterbox TTS offers open-source models for research and development under the MIT licence.
• CosyVoice provides an Apache-2.0 repository with downloadable source code and models.

Verdict

VoiceStudio is a strong fit for creators and developers who want local control over a broad set of voice and audio workflows, particularly when dubbing, transcription, and long-form production sit alongside voice creation. Its local API and user-controlled storage are compelling reasons to choose it; look elsewhere if you need the hosted dashboard in the public build, or if local hardware requirements and separate model licences are a poor fit.