> ## Documentation Index
> Fetch the complete documentation index at: https://docs.superscale.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Speaking characters

> Creator-style videos where an AI character talks straight to camera: the classic UGC testimonial.

A speaking character is an AI creator talking straight to camera: the talking-head UGC testimonial format used for direct-response ads. In the current chat video flow, this is the **UGC Testimonial** Seedance format: English-only, vertical, and generated from a brief, script, and references.

If you need a saved actor, reusable avatar, actor selection, or a more controlled template assembly flow, use the video template/avatar workflow instead of treating chat Seedance as an actor library.

## What makes them work

<CardGroup cols={2}>
  <Card title="Talks to camera">
    A creator delivers your message face-to-camera, with native English speech and lip-synced delivery.
  </Card>

  <Card title="Research-backed scripts">
    The hook and script come from what's winning in your niche, not a generic template.
  </Card>

  <Card title="On-brand by default">
    Built from your [context](/context/overview): your product, your claims, your tone.
  </Card>

  <Card title="Ready to launch">
    Generated as a short 9:16 paid-social video with optional [captions](/features/captions) and end card.
  </Card>
</CardGroup>

## Create one

<Steps>
  <Step title="Describe the creator and angle" icon="message-square">
    For example: *"a friendly 30-something reviewing our protein bar, casual kitchen, problem-solution hook."*
  </Step>

  <Step title="The agent writes and generates" icon="wand-sparkles">
    It scripts the testimonial and produces the English talk-to-camera video with synced delivery.
  </Step>

  <Step title="Iterate to find a winner" icon="copy">
    Try different hooks or creators. See [Iterate on a creative](/features/iterate).
  </Step>
</Steps>

<Note>
  Talk-to-camera testimonials are produced in English with native audio. To reach other languages, use **UGC Lifestyle**, **B-Roll**, or another non-English-capable Seedance format with post-mixed [voiceover](/features/voiceover-lipsync).
</Note>

## What not to use it for

* Real founder impersonation, real employee spokespersons, or fake customer proof.
* Non-English native lip-sync testimonials.
* Exact reusable actors that must appear consistently across many campaigns.
* Horizontal or landscape video output.