Product
Voice Studio is the creative suite of the voice AI platform: cloning, generation, effects, and long-form audio, in the browser, on hosted GPUs. It is part of PeachDesk, not a separate product.
Clone a voice from a sample and manage your voice profiles in the browser. Cloning sits behind a responsible-use policy with a consent notice: clone only voices you have permission to use. The policy is a notice, not an enforcement mechanism, so consent is your responsibility to secure.
Generate multilingual speech, then shape it with effects applied as versioned takes. Every take stays intact, so you can compare reads, keep the best one, and never lose a good version to an edit. Dictate when speaking is faster than typing.
Assemble multi-voice timelines in an editor built for longer audio: several voices, one narrative, produced end to end in the studio.
Voice Studio runs on PeachDesk-hosted GPUs and lives in your hosted account. The agent platform can be fully self-hosted; the studio is not part of the self-hosted package. Voices you create here are made to work with the agents you build on the canvas: two model planes, one platform, one studio.
Prefer to direct the studio from an AI assistant? Voice Studio ships its own voice agent MCP server, so MCP-aware assistants can drive it directly.
Frequently asked questions
No. PeachDesk Voice Studio is part of the PeachDesk platform, not a standalone tool you buy separately. It lives beside the agent builder: you clone or generate a voice in the studio, then put that voice to work in the agents you design on the canvas.
Voice cloning in Voice Studio is governed by a responsible-use policy with a consent notice, and you should only clone voices you have permission to use. The policy is a notice, not a technical enforcement system, so the responsibility for consent sits with the team doing the cloning.
Voice Studio is a hosted capability of the PeachDesk platform: cloning, generation, effects, and Stories run in the browser on PeachDesk-hosted GPUs. The agent platform itself can be fully self-hosted; the studio is not part of the self-hosted package. Teams that self-host the agent platform still use the studio through their hosted account.
Three things. Voices: clone a voice from a sample and manage your voice profiles. Speech: generate multilingual speech, then apply effects as versioned takes so every take stays intact. Stories: assemble multi-voice timelines in an editor built for longer audio, with dictation when you want to speak instead of type.