ALS and motor neurone disease
Keeping a voice, and the things that are free
This page is mostly about other people's programmes, because several of them are free and several of them are better than anything we could offer you. The one thing we ask is that you do not put off the recording.
Read this first if the diagnosis is recent
If you are in the United States or Canada, check whether you qualify for Talk to Me, Goose! through the Live Like Lou Foundation. It is free for registered ALS families, it includes cloud voice cloning, and it runs on iPhone, iPad, Mac, Android and Windows. It was built by David Betts, who has ALS, with Anne Mundell.
Check that before you pay anyone for anything.
When speech starts to change
Two facts shape every decision on this page.
Most people with ALS eventually lose the ability to speak. The figure usually cited by clinicians is around 95 per cent at some point. Bulbar-onset disease compresses the timeline; limb-onset disease usually gives a longer, more gradual decline.
Speaking rate falls before clarity does. This is the finding that drives clinical timing. The rule speech-language pathologists use is to refer for a communication assessment when speaking rate drops to about 100 to 125 words per minute, well before speech sounds obviously affected to a family member. Rapid loss of intelligibility tends to follow once rate has fallen to roughly half of a person's habitual speed.
The practical translation: if you are waiting until the voice sounds wrong, you have already waited too long to record it.
Communication aids are often described as a milestone of decline, and people resist them for exactly that reason. It is worth naming a different frame: banking a voice is not a surrender, it is the same category of act as writing down a family recipe. It is done early precisely because it is worth having later.
Voice banking and message banking
Two different things, often confused, and both worth doing.
Voice banking
You read a set script. A service builds a synthetic voice from the recordings. Afterwards that voice can say anything you type.
Strength: unlimited flexibility.
Weakness: it is a synthesis of you, not you.
Message banking
You record specific words, phrases, sounds and laughs in your own voice. These are stored as audio files and attached to buttons.
Strength: it is genuinely your voice, with your timing and your jokes intact.
Weakness: you can only play back what you recorded.
People who have done both tend to describe the same split: the synthetic voice does the day-to-day work, and the banked recordings carry the things that matter — a grandchild's name, a catchphrase, a specific way of saying I love you. Some people bank a few hundred messages; some bank several thousand.
mymessagebanking.com is a free message banking service run by the Department of Otolaryngology at Boston Children's Hospital, in partnership with Tobii Dynavox. Free for personal use, including recording, storage and download of the audio files. Those files are yours, and they outlive any single app or account.
Practical advice for recording
- Record in short sessions. Fatigue is real and it degrades the recordings.
- Record in the morning if mornings are better.
- Do not chase perfection. A slightly imperfect recording made this month beats a perfect one that never happens.
- Include the unglamorous things: no, stop, that hurts, turn me, I need suction, alongside the sentimental ones.
- Gather old voicemails, videos and voice messages into one folder. Some services can use existing audio.
Apple Personal Voice, free and already on the phone
What it is
Apple Personal Voice creates a synthetic voice that sounds like you, entirely on your own device. It costs nothing, it is built into iOS, and in recent versions it needs about ten recorded phrases rather than the 150 originally required. The training happens on the iPhone, usually overnight while charging, and the result can be shared to your other Apple devices through iCloud with end-to-end encryption.
Where to find it
Settings → Accessibility → Personal Voice. There is no direct link an app can open, so you have to walk that path yourself.
While you are there, look at two switches: Share Across Devices, and Allow Apps to Request to Use. The second one is what lets a communication app speak with your voice.
What it runs on
iPhone 12 and later, and equivalent iPads and Macs.
Two limits worth knowing
- Apps may speak with your Personal Voice, through the speaker or during a call, but they cannot capture, copy or export the audio. That is an Apple rule, and it is a good one. It also means no app can hand your Personal Voice to anyone else.
- A Personal Voice is tied to the Apple Account that made it, and a legacy contact cannot access it after death. Banked message recordings are ordinary audio files and do not have that problem. If the recordings matter to your family, bank messages as well.
The ElevenLabs Impact Program
ElevenLabs is a voice AI company. Its Impact Program gives free licences to people who are losing, or have lost, the ability to speak. This is the single most valuable free thing in this space that most families have never heard of.
- The patient licence is a multi-year professional licence, free to the person. Partner charities have described the value as around $1,200 per year.
- Since October 2025, patients and clinicians can apply directly on the ElevenLabs website, as well as through a partner charity.
- Eligibility is broader than ALS. It covers people over 18 affected by conditions causing permanent voice loss, and has been extended to mouth cancer and laryngectomy through specialist charities, and to conditions including multiple system atrophy and progressive supranuclear palsy through their partner organisations.
- Voice cloning can work from surprisingly little audio, including old voicemails and videos. If speech has already changed, ask anyway.
The ordering trap
Several partner charities warn about this in bold, so we will too. Contact the partner charity before you register on the ElevenLabs website. In the United Kingdom, contact MND Connect at the MND Association first, to get a code to use when registering. Registering first can strand you on a paid path that is awkward to unwind.
In the United States, Bridging Voice handles ALS applications. Their route asks for a letter from your doctor or speech-language pathologist confirming the diagnosis, and they aim to approve within a few days.
The honest limitation: ElevenLabs voices need an internet connection. Voice banking comparison charts from motor neurone disease organisations state it plainly: an online connection is required, though saved history and phrases can be downloaded and used offline. That is fine at home and unreliable in a hospital, a car park, or a rural clinic. Many people run an ElevenLabs voice as their main voice and keep an offline voice as the fallback. That is a sensible arrangement, not a compromise.
Other voice services
- Acapela — a long-established voice banking service, priced at roughly 99 in euros or dollars per year for one platform, or about 999 as a one-time purchase for all platforms, with a free trial. It supports message banking too.
- Your clinic's recommendation — the MND Association voice banking comparison chart is the most useful single document in this area. It is a PDF and it is updated periodically.
Most motor neurone disease organisations decline to endorse a specific service, and recommend discussing the choice with your speech and language professional at the earliest opportunity. That is the right advice, and we are not going to contradict it.
Equipment loan closets
Before buying hardware, ask about borrowing it. Loan programmes lend communication equipment free, and are usually badly advertised.
- ALS Association communication device loan programme.
- Team Gleason — communication technology and equipment for people with ALS.
- ALS Network equipment loan programme.
- ALS Association certified centres and clinics — a multidisciplinary clinic will usually have a speech-language pathologist who does this every week.
- Les Turner ALS Foundation communication guide (PDF).
A device gap worth knowing about
Two Apple features that people assume go together have different hardware requirements, and the gap catches families out when they are choosing a phone or accepting a hand-me-down iPad.
| Feature | Runs on |
|---|---|
| Live Speech (type and speak, including during calls) | Any recent iPhone, iPad or Mac. Free. |
| Personal Voice (your own synthetic voice) | iPhone 12 and later, and equivalent iPads and Macs. Free. |
| Eye tracking, built into iPadOS and iOS | Recent iPhones and iPads with a front TrueDepth camera. Free. |
| Apple Intelligence, which our sentence feature uses | iPhone 15 Pro, 15 Pro Max, and iPhone 16 and later. |
So a person on an iPhone 13 can have their own banked voice, eye tracking and free type-and-speak, and cannot run generated sentences. Our app still works on those devices, with rule-based options instead of generated sentences, and it is free there as it is everywhere.
Head and neck cancer, and a fixed surgery date
Much of this page applies equally to laryngectomy, glossectomy and other head and neck surgery, with one important difference: there is usually a known date.
If surgery is scheduled, the window before it is the whole opportunity. Bank the voice, bank the messages, and set up the app or the free built-in feature before the operation, not in the recovery ward. Free voice programmes have been extended to mouth cancer and laryngectomy through specialist charities, so ask.
After surgery, many people are otherwise physically well, with full use of their hands. If that is the situation, typing speed is the whole problem and free type-and-speak tools may cover it entirely. Start there. See the temporary voice loss section.
Where this app fits, honestly
ALS is not our main audience, and we will not pretend otherwise. There is a free, foundation-funded app for ALS families in the United States and Canada that does starter-words-to-sentences, offers voice cloning, and runs on more platforms than we do. Where that programme covers you, take it.
What our app does differently is narrow and specific:
- It works with the internet off. No connection, no account, no character quota. Sentence generation and speech both happen on the phone.
- Your speech is never billed by the character or the minute. There is no meter.
- Nothing leaves the device. No utterance is transmitted anywhere, because the app has no network permission at all.
- Huge type, four tiles, one button. Designed for the days when fatigue is winning.
If those things matter to you, the app is free, with everything included and no account or card. If they do not, use the free options above with our blessing.
Sources
- Apple Support: Create a Personal Voice — enrolment, on-device training, sharing across devices, and the rule that apps cannot capture Personal Voice audio.
- Apple Newsroom — the reduction to about ten enrolment phrases.
- ElevenLabs Impact Program and the direct application announcement.
- Bridging Voice — the ALS application route in the United States.
- MND Association voice banking services comparison chart (PDF, October 2025) — including the statement that ElevenLabs voices require an online connection.
- mymessagebanking.com — free message banking from Boston Children's Hospital with Tobii Dynavox.
- Live Like Lou — the free Talk to Me, Goose! programme for registered ALS families in the United States and Canada.
- Speaking rate as a referral trigger: Yorkston, Beukelman and colleagues on the Sentence Intelligibility Test; Ball, Beukelman and Pattee on rate decline preceding intelligibility decline.