No blank page.

Voice journal that listens — really.

Speak, don’t type.

Talk for 30 seconds or 10 minutes. Reflecsia transcribes, finds the patterns, and shows you what your week actually felt like.

Free for 10 entries / month. No card to start.
Mood over time
How it works

Three steps. No streaks. No shame.

01

Talk.

Tap once, speak. 30 seconds or 10 minutes — whatever fits the moment.

Recording
02

Read.

Your voice becomes structured text. Audio is deleted right after.

Entry detail
03

See.

Each entry maps emotions, themes, and patterns. Weekly and monthly summaries show what’s changing.

Insights
Why voice

Most journaling apps ask you to write. Reflecsia asks you to talk. People speak about 4× faster than they type — a 5-minute voice entry captures what would take 18 minutes to write down. That’s why most journals get abandoned in week one. And why Reflecsia doesn’t.

faster than typing
5 min
of voice ≈ 18 min of writing
Day 7
when most journals stop. Not this one.
What the AI actually does

What you didn’t notice.

  • Emotions
    Every entry mapped on Plutchik’s wheel + valence / arousal.
  • Themes
    Recurring topics surface across entries automatically.
  • Cognitive patterns
    Gentle CBT-aware reflection (not therapy, not a diagnosis).
  • Reflection prompts
    Questions personalised to what you actually said.
  • Summaries
    Daily, weekly, monthly. The longitudinal view.

Self-observation. The kind that can improve a life.

Entry detail
Privacy & honesty

Audio is deleted right after the transcription completes — it never lives on our servers. Transcripts and analysis are encrypted at rest, scoped to your account by row-level security. We don’t sell data. We don’t train AI on your entries. You can export everything as PDF / Markdown / JSON and delete your account in one tap.

Reflecsia is a self-reflection tool, not a medical device. It helps you notice patterns. If something serious comes up, please talk to a licensed professional.

Read the full privacy policy →

Reflecsia is a tool, not an observer.

Audio deleted on transcribe
Never stored on our servers.
Encrypted at rest
Row-level security, scoped to your account.
No data sold
No tracking pixels, no third-party advertisers.
No training on your entries
Your text doesn’t go into anyone’s model.
Export anytime
PDF, Markdown, or JSON.
Delete in one tap
Account and all data, gone.
Questions
Yes. Audio is deleted immediately after the transcription completes — we never store it on our servers. Only the resulting text and analysis live in your account, encrypted at rest.
No. Your transcripts and reflections are never used to train any model — ours or anyone else’s. They exist only to give insights back to you.
Free for the first 10 entries per month, no card required. Unlimited entries and longer-form summaries are $7.99 / month or $59.99 / year. Cancel any time.
Recording, transcription and AI analysis work in English and Russian today, with more languages rolling out. The app interface follows your phone’s language.
Transcription always runs (it’s how your voice becomes searchable text). The deeper analysis runs in the background — you can ignore the Insights tab and use Reflecsia as a plain voice-to-text journal. The privacy guarantees apply either way.
Yes — available now. The App Store launch is fresh, served via TestFlight while the public listing finishes review. Same features, same privacy guarantees as Android.

Five minutes tonight. A pattern by Sunday.

Free for 10 entries / month. No card to start.