← Architect
Plain answers · Voice

An AI voice journal: speak it, hear it, talk it through

By The Architect team · Last updated

We make Architect. Everything this page says about voice was checked against the app itself on October 23, 2026. The conversation below is a made-up example.

The short answer

Architect is a private AI journal you can use by voice in three ways. You can speak an entry instead of typing it, hear any reply from your mentor read aloud, or open the Sitting and talk it through out loud, back and forth.

Speaking an entry works in 15 dictation languages and is on every plan. The Sitting is in beta.

Speak an entry: Voice Input

Tap Voice Input and say what is on your mind. Your words are transcribed into the entry, and the mentor answers what you said, the same way it answers a typed entry, beside a dated selection of what you wrote before.

  • 15 dictation languages. Choose yours in Settings, Language, Dictation language. The app and the mentor themselves work in English, Turkish and Spanish.
  • No limit on speaking. Voice Input is on every plan, and speaking an entry is not counted against anything; the reply to it counts like any other reply.
  • How the words become text. A voice note is transcribed by a speech-recognition provider. Once it is your entry, it is encrypted on your device before it is stored, like everything you type. When the mentor replies, the text it needs is read by our AI model provider for the moment it takes, as with a typed entry.

Hear a reply: Listen

Under any reply, Listen reads it aloud. In Settings, Your Mentor, Mentor voice, you choose who reads: Auto, which gives each mentor its own voice, or one of eleven voices named after places, each with a sample to hear first.

To make the audio, the reply's text, which can quote what you wrote, goes to a voice provider. On the free plan you have 11 listens in total, one for each of your free replies; the paid plan allows up to 60 a day.

Talk it through: the Sitting

The Sitting is a spoken, hands-free conversation with your mentor: you talk, it answers out loud, you answer back. On a computer it opens from THE SITTING beside History; on a phone, from the round button on the journal.

  • Cutting in. Tap the orb while the mentor is speaking. On a computer you can also just start talking.
  • Pauses. After a long quiet stretch the room pauses, until you tap the orb again.
  • Saved. Each exchange is saved to History, so the conversation stays part of your journal.

It is the closest thing in Architect to an AI mentor voice chat, and it is built around your journal rather than a blank conversation. How the four mentors differ.

What a spoken exchange looks like

A short, made-up example of the Sitting, written out. Nothing in it comes from a real person's journal.

Example · made up

You (spoken): "I keep saying I'll call my dad this week, and then it's Sunday again."

The Architect: "That's the third Sunday you've said it. What happens in your head right before you'd pick up the phone?"

You: "I start planning what he'll say about the job."

The Architect: "So the call you're avoiding isn't with him. It's with the job question. Could the first call be ten minutes about something else?"

The "third Sunday" in a reply like this comes from what the person wrote before, with dates; the mentor only says it when earlier entries hold it. How the memory works.

Honest limits

  • The Sitting is in beta. It works best in a quiet room; background noise can make it hard to tell when you have finished speaking.
  • It needs your microphone. If the browser has blocked it, the app tells you how to allow it.
  • It runs in the browser. There is no native app yet; on a phone, add it to your home screen and it opens full screen like an app.
  • Voice leaves your device to be processed. Speech goes to a speech-recognition provider to become text, and Listen and the Sitting send the reply's text to a voice provider to make the audio. We keep the names of these providers off our website; a list of the service providers we use is available on request at hello@architectapp.ai.

Free and paid

Voice Input, Listen and the Sitting are on every plan. Writing is free, and every entry saves. The free plan includes 11 full mentor replies, then one reply every 7 days. No card is needed. In the Sitting, each spoken reply counts as one reply, the same as a typed follow-up, and as one listen.

The paid plan is $33 a month, $299 a year, or $9.99 a week, with no daily cap on mentor replies and up to 60 listens a day. Every number is on Architect facts.

Frequently asked questions

Can I voice journal with AI in Architect?
Yes. Tap Voice Input and speak an entry; your words are transcribed into the entry and the mentor answers what you said. Dictation works in 15 languages, on every plan.
Can Architect read its replies out loud?
Yes. Listen reads any reply aloud, in Auto (each mentor's own voice) or one of eleven voices you choose in Settings, Your Mentor, Mentor voice. The free plan includes 11 listens; the paid plan allows up to 60 a day.
Can I talk to my AI mentor by voice?
Yes, in the Sitting (beta): a spoken, hands-free conversation with your mentor, back and forth. Tap the orb to cut in. Each exchange is saved to History.
Is voice journaling private?
Your entries are encrypted on your device before they are stored, spoken or typed. To turn speech into text, a voice note goes to a speech-recognition provider, and Listen and the Sitting send the reply's text to a voice provider to make the audio. To write the reply, the text it needs is read by our AI model provider for the moment it takes.
Is voice free?
Voice Input, Listen and the Sitting are on every plan. Writing is free, and every entry saves. The free plan includes 11 full mentor replies, then one reply every 7 days. No card is needed. The paid plan is $33 a month, $299 a year, or $9.99 a week.
Try Architect with one real entry →
Writing free forever · No card · 11 mentor responses to start
Speak your first entry, or type it. Writing is free.