How to Configure Live Speech and Personal Voice Accessibility on iPhone in iOS 18

For individuals at risk of losing their speaking ability due to progressive neurological conditions (such as ALS), or those experiencing temporary vocal strain, synthesized computer speech can often sound robotic and impersonal. In iOS 18, Apple provides transformative speech accessibility tools: Personal Voice and Live Speech. Using on-device machine learning neural networks, iPhone users can record a sample of their natural voice to generate a custom synthetic voice replica, enabling them to communicate naturally during phone calls, FaceTime sessions, and in-person conversations by simply typing.

Understanding Personal Voice vs Live Speech

These two accessibility technologies operate together to provide seamless communication:

• Personal Voice: An on-device generative AI neural model trained on 15 minutes of your recorded speech. The voice training processes entirely locally on the iPhone’s Neural Engine during overnight charging, preserving absolute audio privacy without cloud transmission.
• Live Speech: The user-interface utility summoned via a triple-click of the Side button. It converts typed sentences into spoken audio in real time, routing synthesized speech through the iPhone speakers or into live cellular and FaceTime audio streams.

Recording and Training a Personal Voice on iPhone

To train your personalized synthetic voice replica:

1. Open the Settings app on your iPhone running iOS 18.
2. Scroll down and select Accessibility.
3. Under the Speech section, tap Personal Voice.
4. Tap Create a Personal Voice.
5. Authenticate using Face ID, Touch ID, or your device passcode.
6. Find a quiet indoor room with minimal acoustic reverberation. Hold the iPhone approximately 6 to 12 inches from your mouth.
7. Follow the on-screen prompts to read 150 short randomized phrases aloud. Each phrase takes a few seconds, with total recording time averaging 15 minutes.
8. Once complete, connect your iPhone to power and lock the screen. The iPhone processes and synthesizes the neural model overnight while connected to Wi-Fi and charging.

Enabling and Customizing Live Speech

Once your Personal Voice finishes generating (or if you prefer using Apple’s high-fidelity system voices):

1. In Settings > Accessibility, select Live Speech.
2. Toggle Live Speech to On.
3. Tap Voices. Under the English category, select your generated Personal Voice (or choose expressive Siri voices).
4. Return to the Live Speech menu and tap Favorite Phrases.
5. Add common daily conversational responses (e.g., “Hello, my name is Alex,” or “Could you please repeat that?”) to enable instantaneous one-tap spoken delivery.

Invoking Live Speech During Calls and In-Person Discussions

To communicate using your synthesized voice:

1. During a Phone or FaceTime Call: Triple-click the Side button (or tap the Live Speech icon in Control Center). A discreet text bar appears on the screen.
2. Type your response using the keyboard or tap the Phrases shortcut icon.
3. Tap Send on the virtual keyboard. The iPhone speaks your sentence aloud across the call audio stream using your custom voice timbre.
4. During In-Person Conversations: Summon Live Speech at any time while the phone is unlocked; audio plays directly through the iPhone stereo speakers with crystal clarity.

Securing and Syncing Across Apple Ecosystem Devices

Personal Voice models are encrypted with end-to-end security. If you enable Share across devices in Settings > Accessibility > Personal Voice, your trained voice replica synchronizes securely to your iPad and Mac via iCloud Keychain, allowing you to use Live Speech across all your devices without re-recording audio samples.