Changelog

Workspaces are now a first-class part of SonicVox. Invite teammates by email — they don't need an account first — switch between your personal workspace and any team you belong to, and see at a glance which one you're working in.

Everything you create belongs to the workspace it was made in: voices, projects, history, generated files and folders. Share a Library folder and everything inside it comes with it, including anything you add later.

A team seat now buys usage, not just access. Members of a workspace draw from the owner's balance instead of needing their own plan, and the balance shown to a member is the balance they can actually spend.

Owners keep full visibility: usage is still attributed per member in your billing history.

Every v1 API endpoint now returns errors in one consistent shape and includes a request ID you can quote to support. The developer docs have been rewritten around code you can paste and run, with the cost of each call stated up front.

Drop audio files anywhere on the Library page to import them — no file picker, no upload dialog.

Audio you generate now carries a standard AI-generated marker inside the file itself, not just in our records. It travels with the file wherever it goes, which is what disclosure rules in a growing number of markets ask for.

Dubbing, separation and other long-running jobs now notify you when they finish or fail, so you can close the tab and come back later instead of watching a progress bar.

Speech Separation is no longer limited to two voices. Upload a recording with several people talking — including passages where they talk over each other — and get one clean track per speaker.

The upload screen now states the speaker counts we have actually measured, so you know what to expect before you spend credits.