Status First
The Voice Studio library, the built-in voices that stand in for recording sessions, is a frozen design, and as of October 2026 narration on x-rush.ai is not live yet: no production voice integration, no launch date to print. The Audio module's live feature is Music Gen. This page documents the design as specified. If you are reading it after launch, the Voice Studio page carries the current status.
The Voice Library
The library holds two kinds of voices, both scoped to the Project you are working in:
- Built-in voices. A first batch of voices licensed for product use and passed through quality review. The launch set is deliberately curated rather than large; no "hundreds of voices" promise appears in the spec, and this page won't invent one.
- Private clones. Consent-gated voices made from your own samples, private to the Project (see Voice Cloning).
You can search the library and filter by language, accent, scene, and what each voice can actually do. Which languages ship in the first batch is settled by evaluation before launch, not by a number on this page; each voice shows the languages and accents it really supports.
What You Can Control
Controls exist only where a voice truly supports them, and the interface shows exactly that set per voice:
- Emotion and style, where the voice supports them
- Speed — pacing for the content
- Pauses — dramatic or conversational breaks, where supported
- Pronunciation and pitch, for voices that expose them
One deliberate absence: there is no universal per-word emphasis control. If a voice doesn't support a control, switching to it normalizes the old parameter to defaults rather than silently keeping an effect that can't render.
What It's For
The design targets the usual narration work: YouTube and course narration, podcast episodes, audiobook production, corporate training, and serialized content where re-recording the same voice every week is the real cost. Narration runs as whole-body text against one voice per generation; multi-speaker dialogue and sentence-level surgery are not part of the first design. For music, the live alternative today is Music Gen; for the voice work above, the honest answer is wait for launch.