Losing the ability to speak due to medical conditions such as ALS or vocal cord damage can be life-altering. Apple introduced Personal Voice iPhone integration within iOS to enable users at risk of speech loss to synthesise a digital replica of their own voice. Integrated tightly with Live Speech, this accessibility feature allows you to type what you want to say during FaceTime calls, phone conversations, and in-person discussions while sounding like yourself.
What Is Personal Voice and How Does It Work?
Personal Voice is an accessibility tool powered by on-device machine learning. By recording approximately 150 short phrases over a 15-minute guided training session, your iPhone models your unique vocal characteristics, inflection, and timbre. To safeguard your privacy, all training data and model synthesis are processed entirely on-device using the Apple Neural Engine, ensuring your raw voice recordings are never uploaded to external cloud servers.
Requirements for Setting Up Personal Voice
Before beginning the voice creation process, ensure your hardware and software meet the following prerequisites:
- An iPhone running iOS 17 or later equipped with an A14 Bionic chip or newer (iPhone 12 series and newer).
- Face ID, Touch ID, or a device passcode enabled on the device.
- Two-factor authentication configured on your Apple Account.
- A quiet environment with minimal ambient noise, echo, or air conditioning rumble.
- A stable Wi-Fi connection and an iPhone charger, as background voice generation requires the phone to be locked and plugged in overnight.
How to Record and Create Your Personal Voice
Generating your voice replica requires following a structured on-screen reading session:
- Open the Settings app on your iPhone.
- Scroll down and tap Accessibility.
- Under the Speech section, select Personal Voice.
- Tap Create a Personal Voice and authenticate using Face ID or your device passcode.
- Follow the introductory guidance and tap Continue.
- Assign a recognisable name to your voice profile (such as your first name).
- Tap Record to perform a quick audio quality test. Read the test prompt aloud at your natural conversational volume.
- Once your ambient sound level is verified, proceed through the 150 prompt cards, speaking each phrase clearly and naturally.
If you stumble over a word or background noise interrupts you, tap Listen to review the clip or Record Again to re-record that specific phrase.
Processing the Voice Replica Overnight
Once you finish reading all 150 prompts, the initial recording stage is complete. However, training the neural synthesis model requires substantial computational power:
- Connect your iPhone to its charger.
- Lock the device screen. Processing will pause if the device is unlocked or in active use.
- Ensure the iPhone remains connected to power throughout the night.
Depending on your iPhone model and battery temperature management, the synthesis process typically takes between 4 and 10 hours. Your iPhone will display a notification once your Personal Voice is ready for use.
How to Use Personal Voice with Live Speech
Once generated, your voice profile links directly with Apple’s Live Speech feature:
- Navigate to Settings > Accessibility > Live Speech.
- Toggle the switch next to Live Speech to the On position.
- Tap Voices, select Personal Voice at the top of the menu, and choose your newly generated profile.
- Triple-click the side button (or Home button on supported hardware) to summon the Live Speech dialogue box from any app.
- Type your sentence into the text box and tap Send. Your iPhone will vocalise the text using your synthesised voice through the speaker or during active phone calls.
Syncing Personal Voice Across Apple Devices
If you own an iPad or Mac compatible with Personal Voice, you can share your voice profile seamlessly without re-recording the 150 phrases. In Settings > Accessibility > Personal Voice, enable Share Across Devices. Your trained model will sync securely across your iCloud account with end-to-end encryption.