Your voice is part of who you are. With Voice Banking, you can create a personal digital voice that sounds just like you — or preserve the voice of a loved one — and use it everywhere you communicate in Flexspeak.
What is voice banking?
Voice banking builds a personal text-to-speech voice from a few minutes of audio. It is powered by ElevenLabs, works in 30+ languages, and is free during the beta.
Before you begin
Because a cloned voice can identify a real person, we ask for clear consent before creating one:
- You must be 18 or older to create a voice, and the voice being cloned must belong to an adult (18+). ElevenLabs doesn't allow voice clones of anyone under 18, even with a parent's permission.
- You can only clone your own voice, or a voice you have the speaker's written permission to clone.
- Your recordings are sent to ElevenLabs only to build the voice model — Flexspeak does not keep your recordings. The voice model is stored with ElevenLabs until you delete it.
Voice cloning is covered by our Privacy Policy and AI Terms.
Option 1: Record your voice (mobile app)
The fastest way to create your voice — it takes about 2 minutes.
- Tap the Settings icon.
- Tap Speech & Language.
- Tap Set up Voice Banking.
- Review the consent step and check the boxes to agree.
- Read the 5 short sentences out loud in a quiet place. Want an even closer match? Record up to 5 extra sentences.
- Tap Create my voice. Your voice is usually ready in under a minute.
Option 2: Upload recordings
Already have recordings? Upload voice memos, voicemail exports, or audio from videos instead of recording live. This is especially helpful for preserving the voice of a loved one.
- Open the voice clone wizard (steps 1–3 above). On the Flexspeak web app, uploading is the default. On mobile, tap Upload recordings instead.
- Review the consent step, including the confirmation that the recordings are your own voice or that you have written permission from the speaker.
- Tap Choose audio files and add 1–10 recordings (m4a, mp3, wav, webm, ogg, flac, or aac — up to 10 MB per file).
- Tap Create voice.
Tips for the best-sounding voice
- 1–5 minutes of clear speech in total is the sweet spot — more than that doesn't improve quality.
- One speaker only — no music, TV, or background chatter.
- Keep recordings consistent — clips recorded in the same place on the same device work much better than a mix of very different recordings.
- Natural, expressive speech clones better than flat reading.
Using your cloned voice
Right after creation, choose where the voice should be used — all supported languages (recommended) or just your current language.
You can also switch voices any time:
- Tap the Settings icon.
- Tap Speech & Language, then Languages & Voices.
- Pick your voice under Cloud Voices.
Renaming or deleting a voice
- Tap the Settings icon.
- Tap Speech & Language.
- Tap Banked Voices to preview, rename, or delete any of your voices.
Deleting is permanent. When you delete a cloned voice, the voice model is also deleted from ElevenLabs and cannot be recovered. Deleting your Flexspeak account deletes all of your cloned voices as well.