Bank your voice in ten seconds with Voice Back
Voice Back is a voice banking app that clones your voice from about ten seconds of audio — including an old voicemail, home video, or interview. It is designed for people facing ALS, MND, cancer, surgery, stroke, and other conditions that can take away speech, and it still works even after speech has already changed.
What is voice banking?
Voice banking is the process of recording a large set of scripted sentences while a person can still speak clearly, so software can later build a synthetic version of that voice for use in an AAC app or speech generating device. Most programs ask for 50 to 200 sentences; some ask for more than 1,600.
Voice banking works well when there is time and energy to complete it. The hard part is that ALS, cancer treatment, and surgery rarely give people that runway, and fatigue or slurring during recording degrades the finished voice.
What is message banking?
Message banking is the practice of recording specific phrases in a person's own natural voice and saving those exact audio files to play back later. Message banking preserves real intonation and laughter, which families treasure, but its limit is that it can only replay what was already recorded — it cannot say a new sentence typed tomorrow.
How does Voice Back's AI voice cloning work?
Voice Back uses AI voice cloning to build a digital copy of your voice from a short audio sample. Instead of asking you to record hundreds of scripted sentences, the model analyzes the unique characteristics of your voice — pitch, tone, cadence, and pronunciation — from a single clip of about ten seconds.
Once the clone is created, you can type anything into the app and hear it spoken in your voice. You can organize phrases into tabs, customize buttons for common needs, adjust speaking speed, and even add custom pronunciations for names, places, and family nicknames.
Because the clone can be made from older recordings, Voice Back is often used as an ALS voice banking option when traditional banking is no longer possible. A wedding toast saved on a phone, a voicemail left for a friend, or an old interview can all become the source of a restored voice.
How is a 10-second voice clone different from voice banking?
A 10-second voice clone uses AI to learn the timbre and cadence of a voice from a short sample rather than from a full recorded script, which is how Voice Back builds its clone. Ten seconds of clean audio is enough to generate unlimited new speech in that voice. If speech has already changed, an older clip — a voicemail, a wedding toast, a home video — can be uploaded and cloned from instead.
You can also teach it names, places, and family nicknames with custom pronunciations, and speak through a customizable board with tabs and buttons.
How does Voice Back compare to traditional voice banking?
| Voice Back | Traditional voice banking | |
|---|---|---|
| Recording required | One clip of about 10 seconds | 50–200+ scripted sentences, sometimes 1,600+ |
| Time commitment | Minutes | Hours, usually across multiple sessions |
| Can use old audio | Yes — voicemails, videos, interviews | No |
| Works after speech loss | Yes, if any recording of the voice exists | No |
| Generates new sentences | Yes, anything you type | Voice banking yes, message banking no |
| Where it runs | iPhone, iPad, and Android | Often desktop software or clinic equipment |
| Cost | $2.99/month or $23.99 lifetime | Frequently $100–$600+ |
Is Voice Back the right voice banking app for you?
The right choice depends on time and energy: message banking is still worth doing when there is time, energy, and a clinician guiding the process, since those recordings are irreplaceable, while Voice Back is the option when time is short, recording sessions are exhausting, or speech has already changed. Many families do both.
See pricing or read how Voice Back is used by people living with ALS.
Last updated: