Biteable
All articles

AI voiceovers: text-to-speech and ElevenLabs voices

Generate natural-sounding narration from a script with Biteable's text-to-speech AI voices — no recording needed. You can use Biteable's built-in voices or bring a voice from ElevenLabs.

Generate an AI voiceover

  1. In your project, click Voice-over in the content bar (or the microphone icon in the timeline) and open the Text to Speech tab.
  2. Enter or paste your script. There's a per-clip character limit (about 1,000 characters, roughly 200 words) — split longer narration across multiple clips. The panel also includes built-in translation and an AI text assistant for reworking your script, plus a voice-speed toggle.
  3. Choose a voice on the Biteable voices tab — click Preview to hear a sample of each.
  4. Click Generate voiceover, preview the result, then click Add to timeline.

The panel shows how many voice-over clips you have remaining when you approach your generation limit.

Replace or update an AI voiceover

Select the voiceover clip on the timeline and choose Replace voice-over. Edit the text or pick a different voice, click Update voiceover, then Add to timeline to swap in the new version.

Use an ElevenLabs voice

If none of Biteable's built-in voices are right, you can bring in a voice from ElevenLabs' voice library. You'll need a free ElevenLabs account. Only the voice ID crosses over — the audio is still generated in Biteable and counts toward your Biteable voice-over generations.

Keep your own record of any voice ID you want to reuse. Biteable doesn't store ElevenLabs voice IDs — there's no list of voices you've used before, and the field is empty each time you open the panel. You paste the ID in every time you generate a voiceover with that voice, so save the ones you like somewhere you can get at them. The easiest way is to add the voice to My Voices in ElevenLabs, so you can always go back and copy the ID again.

1. Find a voice in ElevenLabs and copy its ID

  1. Sign in at elevenlabs.io and go to Voices.

  2. On the Explore tab, find a voice you like. Search by name or keyword, or narrow the list with the Language and Accent dropdowns and the category chips (Conversational, Narration, Characters, Social Media, Educational). Filters has more options, including gender, age, and quality.

    The ElevenLabs Explore tab, showing the voice search box, the Language and Accent filters, the category chips, and a grid of trending voices.

  3. Click the play button on a voice to hear a sample.

  4. On the voice you want, open the More actions menu (the three dots) and choose Copy voice ID. You'll get a short string of letters and numbers, something like 21m00Tcm4TlvDq8ikWAM.

    An ElevenLabs voice row with the three-dot More actions menu open, showing Copy voice link, Copy voice ID, and View similar.

Voices you've created or saved to your account live on the My Voices tab, which has the same Copy voice ID option. You don't need to save a library voice to My Voices first — you can copy the ID straight from Explore — but saving it there makes the ID easy to find again later.

2. Add the voice to your Biteable project

  1. In your project, open Voice-over → Text to Speech.
  2. Under choose a voice, click the Add ElevenLabs voice tab.
  3. Paste the ID into the Paste ElevenLabs Voice ID To Add field.
  4. Enter your script, click Generate voiceover, preview the result, then click Add to timeline.

The voice is loaded at the moment you generate, so there's no preview of it in Biteable beforehand. To use the same voice on another clip, paste its ID in again.

Not every ElevenLabs voice ID works through the API Biteable uses. An unsupported voice can fail with an error like "there are invalid characters in your script" even when your text is fine. If that happens, use a different voice ID — or generate the audio in ElevenLabs directly, export it, and upload the file to your voiceover timeline.

Customize narration with SSML (Amazon voices only)

Some of Biteable's built-in voices are Amazon-based, and those support SSML tags for fine-tuning narration — pauses, speed, and volume. Type the tags directly into your script.

Most built-in voices are ElevenLabs voices, and they ignore SSML tags. To change the pace of an ElevenLabs voice, use the voice-speed toggle in the Text to Speech panel instead.

Telling them apart: in the voice picker, Amazon voices usually have a coloured or pastel avatar background, and ElevenLabs voices have a white or off-white one. The named list further down is the definitive one — and if you add a tag and it makes no difference to the audio, that voice doesn't support it, so try another.

Add a pause

Use <break/> for a short default pause, or <break time="2s"/> to set a specific length.

Example: Hi there. Welcome to Biteable. <break/> Let's take a deep breath before we get started. <break time="2s"/> Ok. That was nice.

Adjust speed

Wrap text in <prosody rate="…">…</prosody> to change the speaking rate. Options: x-slow, slow, medium, fast, x-fast.

Example: I'll count up fast, then down slowly, <prosody rate="x-slow"> one two three four five </prosody> <prosody rate="x-fast"> five four three two one </prosody>

Adjust volume

Wrap text in <prosody volume="…">…</prosody> to change loudness. Options: silent, x-soft, soft, medium, loud, x-loud.

Example: This is a normal volume. <prosody volume="x-loud"> This is extra loud </prosody> <prosody volume="x-soft"> and this is extra soft </prosody>

Which Amazon voices support which controls

Amazon voices come in four types: Generative (most expressive), Long-Form (most natural for longer content), Neural (natural and human-like), and Standard. Not every control works on every voice:

  • Becky (Long-Form): Pause, Speed, Volume
  • Ruth (Generative): Pause
  • Chelsea (Neural): Pause, Speed, Volume
  • Salli (Neural): Pause, Speed, Volume
  • Gibson (Generative): Pause
  • Stephen (Generative): Pause
  • Matthew (Neural): Pause, Speed, Volume
  • Joelle (Long-Form): Pause, Speed, Volume
  • Danielle (Neural): Pause, Speed, Volume
  • Joanna (Generative): Pause
  • Gregory (Neural): Pause, Speed, Volume
  • Willow (Generative): Pause
  • Amy (Neural): Pause, Speed, Volume
  • Arthur (Neural): Pause, Speed, Volume
  • Emma (Neural): Pause, Speed, Volume
  • Brian (Neural): Pause, Speed, Volume
  • Leah (Generative): Pause, Speed, Volume

Any built-in voice not on this list is an ElevenLabs voice, so SSML tags won't apply to it.

Troubleshooting

  • "Whoops, couldn't generate the audio": the Amazon-based voices can't handle certain punctuation — the ampersand (&) is the usual culprit. Replace "&" with "and", or switch to a different voice. If every voice fails, clear your browser cache and hard-refresh.
  • A pause/speed/volume tag does nothing: SSML tags only work on the Amazon-based voices, and not every Amazon voice supports every control — see Which Amazon voices support which controls above. On an ElevenLabs voice, use the voice-speed toggle instead.
  • The ElevenLabs voice ID field is empty when you come back to it: that's expected — Biteable doesn't keep a record of voice IDs you've used. Paste the ID in again each time you generate.

Plan availability

Text-to-speech is included on Premium and Business plans, and during your free trial (trials include Premium features). On Pro, you'll see an upgrade prompt — text-to-speech tracks you created during a trial or on Premium stay in your projects, but they can't be edited or regenerated without a plan that includes the feature. Uploading your own voiceover audio works on every plan — see Add, record, and adjust voiceovers.

Related

Still need help?