Help Center

Voice Input & Read Aloud

Speak your messages and listen to AI responses with built-in voice features.

Overview

Sometimes typing does not capture what you want to say. Onsen supports voice in both directions — you can speak your messages using voice input, and you can listen to the AI guide's responses using Read Aloud. Both work across chats and guided experiences.

Why voice makes a difference
Speaking lowers the cognitive load when you're stressed or overwhelmed — sometimes the words come easier when you don't have to type.
Hearing responses read aloud creates a more human, conversational experience that feels like real support.

Voice Input

Instead of typing, tap the microphone button to record your message. You will find it at the right side of the input bar — in a free-form chat it sits next to the text field, and in a guided experience it appears alongside the send button.

Microphone button in a free-form chat
Microphone button in a free-form chat
Microphone button in a guided experience
Microphone button in a guided experience

Speak naturally — you will see a waveform and a timer while recording. When you are done, tap the tick. Onsen writes your words into the message box as it goes, with a bar showing its progress, so you can review and edit them before you tap to send. Names from your journal are spelled the way you write them.

If a recording cannot be written up (for example, the connection drops), it is kept on your phone: tap Try again, or Discard it. Recordings are deleted from your phone once they are written up.

To cancel a recording, tap the cancel button to the left of the waveform. Early in a recording, this throws it away. Once your words have started to appear in the message box, they stay there: only what you said after the last words that appeared is left out.

Long Recordings

There is no time limit on a recording: talk for as long as you like. On a long recording, your words start appearing in the message box while you are still talking, so the wait after you tap the tick stays short.

A message holds up to 10,000 characters, which is roughly 15 minutes of speech. As you get close, a note above the recording controls says Your message is nearly full. Tap the tick to finish, then carry on in the next one. If you reach the limit, the recording finishes by itself, just as if you had tapped the tick. Send the message, then tap again to keep going.

The note when your message is nearly full
The note when your message is nearly full

If part of a long recording cannot be written up (for example, the connection drops for a moment), Onsen says Part of your recording could not be written and how much is missing. That part is kept on your phone: tap Try again to add it in its place, or Leave it out to keep the rest without it.

Recording a voice message
Recording a voice message
Voice recording in a guided experience
Voice recording in a guided experience

Combining Voice and Text

Voice transcriptions are appended to any text you have already typed, separated by a line break. You can type part of a message and dictate the rest, or record multiple voice segments in a row — whatever feels natural.

Microphone Permissions

Onsen needs microphone access to record. If you have not granted it yet, a system prompt will ask for permission the first time. Voice transcription also requires an internet connection. If you have reached your transcription usage limit, you will see a notification — check your usage in Settings.

Read Aloud

Want to listen instead of read? Tap the speaker icon below any AI response to hear it read aloud. Read Aloud works on any AI message, including older ones — scroll up and tap the speaker icon on any past response.

The Player

A player appears at the top of the screen while Onsen reads. The top row is the playback bar, with the time so far on the left and the length on the right. The bottom row has the controls:

  • Playback bar — tap or drag it to jump to any point. While the voice is still loading, the length shows with a ~ because it is an estimate.
  • 1× — tap to change the speed: 1×, 1.25×, 1.5× and 0.75×. Onsen remembers your speed for next time.
  • 10 / 10 — skip back or forward 10 seconds, even while paused or while the voice is still loading.
  • Play / Pause — start or pause playback.
  • Close — stop and dismiss the player.

Every control works with VoiceOver, and the playback bar can be adjusted with a swipe up or down.

The Read Aloud player
The Read Aloud player

Follow Along

While Onsen reads, the part being read is highlighted: each paragraph, heading, list item and quote in turn. The screen scrolls to keep it in view, so you can read along or glance back at any point. If you scroll away to look at something else, following pauses; tap a part or press play to pick it up again.

Want to hear something again, or skip ahead? Tap any paragraph or list item and Onsen plays from there. Tapping works while the player is open, playing or paused.

The paragraph being read is highlighted
The paragraph being read is highlighted

Auto-Speech

You can have every new AI message read aloud automatically — no need to tap the speaker icon each time. To enable it, tap the menu in the top-right corner of the chat screen and turn on the Read Aloud toggle. Each new response will be read aloud as soon as it finishes generating. Turn the toggle off to go back to manual playback.

Some experiences — like meditations and relaxation exercises — start Read Aloud automatically regardless of the toggle, so you can close your eyes and just listen. You can still stop playback at any time by tapping close on the playback bar.

Read Aloud and Expressive Voice in the experience menu
Read Aloud and Expressive Voice in the experience menu

Expressive Voice

Read Aloud supports an Expressive Voice toggle in the same experience menu. When enabled, spoken responses include emotional intonation, natural pacing, and richer expression — making conversations feel more like talking to a real person. Expressive Voice works with both Standard and HD voices. HD voices are a Premium feature that offer higher-quality audio with even more natural delivery.

Tips

  • Speak at a normal pace — in a quiet environment for the best transcription.
  • Mix and match — voice and text work together in the same conversation. Use whichever feels right.
  • Listen on the go — Read Aloud is great for revisiting conversations during a walk or commute.