Skip to content
Truly Imagined Truly Imagined Docs

Generative Studio

The Generative Studio is where you turn a script into finished voice and video using consented, licensed likenesses. It combines a generator and a non-destructive editor: create an asset from a prompt, then trim, arrange, layer and render it into something polished — without ever altering your original takes.

What you can make

Section titled “What you can make”
  • Voice — spoken audio from a script, in a chosen actor’s voice.
  • Talking-head video — a lip-synced performance to camera.
  • Combined — voice and video generated together.
  • Compositions — multi-clip audio mixes and video scenes built from any assets you own, across every project.

Your models

Section titled “Your models”

If you’re an actor generating as yourself, you first build your model — the reference data your likeness is generated from. There are three, each for a different kind of output:

  • Voice Model — audio only (voiceover, audiobooks, podcasts).
  • Video Model — film & cinema (talking-head video).
  • 3D Model — gaming & animation (a Gaussian-splat body replica).

A guided capture kit walks you through recording (or uploading) each one, with an optional expression set to widen your range. See Building your model. Creators generating with a licensed or the founder likeness don’t build a model — the likeness is provided by the licence.

How it flows

Section titled “How it flows”
  1. Create a project and choose the model (a licensed actor, or the platform default).
  2. Generate assets from prompts — each one is saved as a development draft.
  3. Edit and compose in the audio or video editor — trim, sequence, layer and mix.
  4. Render a composition into a new asset, keeping every earlier version.
  5. Promote the takes you’re happy with from development to production.

Development vs production

Section titled “Development vs production”

Everything you generate starts as a development asset — watermarked, metered, and perfect for iterating. When a take is final, you promote it to production. This keeps experimentation cheap and your finished library clean.

Tokens at a glance

Section titled “Tokens at a glance”

Generation and rendering are metered in tokens — roughly one token per second of voice and ten per second of video. Drafts in development are inexpensive; the full charge applies when you promote. See Tokens & billing for the detail.

Section titled “Related”