Choose a Voice. Control the Delivery.

Use the default voice, custom voice, uploaded audio, or ElevenLabs—then adjust emotions to match your message.

Pick the workflow that fits your project, then use Emotions Control to make the delivery feel calm, happy, surprised, and more—so your message lands the way you intended.

At a glance

Choose Your Voice Workflow

Four ways to add speech—plus emotions to control the delivery.

Basic/Default voice

  • Best for: fastest setup
  • Language: English
  • You provide: script text
  • Tip: pair with Emotions Control for tone

Custom voice

  • Best for: consistent brand voice
  • You provide: voice sample (min. 15s)
  • Output: generated speech in that voice

Upload voice file

  • Best for: exact performance
  • You provide: pre-recorded audio file
  • Use when: you already have the final delivery

ElevenLabs voices

  • Best for: voice variety + multilingual
  • You provide: ElevenLabs API token + selected voice
  • Languages: 25+ (via ElevenLabs)
voices graph

Option 1: Basic/Default voice (fastest)

Basic/Default voice is the quickest way to generate a video.

  • How it works: Type your script — adjust voice settings — generate.
  • Language: Currently English.

Voice settings (Emotions Control)

Want the same script to feel more upbeat, calmer, more urgent, or more reflective? Use Emotions Control.

You can adjust:

  • Temperature (conservative — creative)
  • Happy
  • Sad
  • Afraid
  • Disgusted
  • Melancholic
  • Surprised
  • Calm

Want examples + templates? Read: Emotions Control Voice Settings →

Tip: For the most natural result, start with one primary emotion, then keep the rest low.

Option 2: Custom voice (your voice sample)

Custom voice is for when you want your videos to sound like a specific person (you, a founder, a spokesperson, a brand voice).

  • How it works: Upload a voice sample — LipSynthesis uses it as the voice for generation.
  • Best for: Brand consistency across many videos.
  • Minimum length: At least 15 seconds.

Option 3: Upload voice file (pre-recorded audio)

This option is different from Custom voice.

  • What you upload: An audio file (MP3/WAV) that contains the exact spoken text you want in the avatar video.
  • Best for: When you already have the perfect delivery (timing, emphasis, pauses) and want the avatar to match it.

Option 4: Use voices from your ElevenLabs account

If you want more voice variety (and multilingual support), you can use voices from ElevenLabs.

  • How it works: Add your ElevenLabs API token in settings — choose a voice from your ElevenLabs account — generate.
  • Languages: ElevenLabs supports 25+ languages (availability depends on your ElevenLabs plan and selected voice).

Choose the Right Option

Choose Basic Voice
for speed

Choose Custom Voice
for a consistent brand voice

Choose Upload Voice File
to use audio you already love

Choose ElevenLabs
for 25+ languages and more voice options

Frequently Asked Questions

Does LipSynthesis support 25+ languages?

If you connect ElevenLabs, you can access voices that support 25+ languages (based on ElevenLabs). The Basic/Default voice option is currently English.


What’s the difference between Custom voice and Upload voice file?

  • Custom voice = a voice sample used to generate new speech.
  • Upload voice file = a pre-recorded audio track that already contains the exact text you want spoken.

What is Temperature?

Temperature controls how conservative vs creative the delivery feels. Conservative is steadier; creative is more expressive.


Can I control emotions?

Yes. Emotions Control lets you shape delivery using sliders like Happy, Calm, Surprised, Melancholic, and more.


How do I use Emotions Control?

Start with one primary emotion (like Calm, Happy, or Surprised), keep the others low, and generate 2–3 variations to compare.

For step-by-step examples and copy-paste templates, read: Emotions Control Voice Settings →

Ready to try it?

Pick an avatar, choose a voice option, and generate a few variations. You’ll hear the difference immediately.

* No credit card required
ai avatar generator interface