Built around three rituals.
An adaptive prompt, lined paper, room for a single honest line. The brain notices what you point at — this is the pointing.
Tuck favourite entries into a jar. On a hard day, pull one at random — past you wrote it for present you.
Research-backed exercises and multi-week journeys — cognitive reframing, letters to your future self, the 5 Fs weekly review.
Your journal, read back to you.
Most journals store what you write. Kansha reads across months of it and tells you what you can’t see from the inside — who keeps appearing, when you’re at your best, and what’s quietly stopped coming up.
Notice this
“Your best weeks all start with a Sunday walk. The three weeks you skipped it are the three you described as ‘flat’.”
The people in your entries
Appears on almost every good day.
Always after a phone call.
Nothing since early May.
What’s gone quiet
Things you used to write about that haven’t appeared in a while. Not an alarm — just a noticing.
Your time signatureCOMPUTED
When you write, and when you write best. No AI call — this one updates the moment you save.
Colour = average mood in that slot · number = entries written
Illustrative output. Yours is built entirely from your own entries.
Built on work that has been tested.
Every exercise in Kansha comes from published psychology, not from a wellness blog. The papers are named so you can go and read them.
“People who wrote weekly about things they were grateful for reported higher wellbeing and fewer physical complaints after ten weeks.”
Emmons & McCullough (2003)
Journal of Personality and Social Psychology
Read the paper →
Used in Kansha for: The daily journal
“A single gratitude-letter exercise produced measurable increases in happiness that were still present a month later.”
Seligman, Steen, Park & Peterson (2005)
American Psychologist
Read the paper →
Used in Kansha for: Gratitude Letter
“Writing about a future in which things had gone well raised subjective wellbeing and self-rated health.”
King (2001)
Personality & Social Psychology Bulletin
Read the paper →
Used in Kansha for: Best Possible Self
“Feeling connected to your future self improves long-term decisions — people who feel it save more and plan further ahead.”
Hershfield (2011)
Annals of the New York Academy of Sciences
Read the paper →
Used in Kansha for: Letter to your future self
Worth holding these lightly. Most of this research uses non-clinical samples, effect sizes against active control conditions are modest, and the Emmons & McCullough finding quoted above comes from their weekly writing condition, where Kansha suggests a daily habit. Gratitude journalling is well evidenced as a small, real help — not as a treatment, and not as a replacement for therapy or medication.
The engine
The analysis runs on Claude.
Anthropic’s Claude does the reading. Nothing is analysed automatically — you press a button, and before it runs you see exactly which entries are about to be sent. Entries you’ve locked are never included, and nothing you write is used to train any model.
Or bring your own key
Add your own Anthropic API key in Settings and the analysis runs on your account, at your cost, as often as you like — no limits from us. console.anthropic.com ↗
The free tier is the whole journal.
Kansha is in beta, so every Pro feature is free for everyone right now.
- ✓Daily entries, jar & practice
- ✓Mood tracking + days written
- ✓3 deep reflection questions
- ✓1 AI insight / month
- ✓Chrome new tab extension
- ✦Everything in Free
- ✦Weekly AI patterns (24h)
- ✦All reflection questions
- ✦Food & energy diary
- ✦Year goals tracker
- ✦Full history & exports
- ✦Voice transcripts
🎉 All Pro features are free during beta.
- ✓Everything in Free
- ✓Use your Anthropic API key
- ✓Unlimited AI insights
- ✓No subscription needed
Before you start
The things people ask first.
Is it really free?
The journal is free forever — writing, mood, the jar, search, your whole history. That isn't a trial.
There's a Pro tier at £2.99/month for more frequent AI analysis. Everything in Pro is free during the beta, so right now there's no difference.
I've tried gratitude apps before and given up. Why would this be different?
Honestly? It might not be. But most gratitude apps fail for the same two reasons, and both are design choices rather than facts of life.
The first is streaks. Miss a day, break the chain, feel like a failure, delete the app. Kansha has no streak counter anywhere. There is nothing to protect and nothing to lose.
The second is that listing three things gets boring by week two. So there's more here than a list — a jar of kept moments you can open on a bad day, structured exercises from CBT and positive psychology, and an analysis that reads your entries back to you.
Can you read my journal?
No. Entries are protected by row-level security at the database, which means queries can only ever return rows belonging to the signed-in account. There's no admin screen for browsing user entries, and I've never read anyone's.
If you'd like the longer technical version, it's on the Trust page.
What does the AI actually do?
It reads your recent entries and looks for things that are hard to notice from inside your own life. Who appears most often and how they show up. Which times of week your mood tends to be highest. What you used to write about that's quietly disappeared. One line it most wants you to see.
It's the part of Kansha people find most surprising, and it's the reason most keep using it.
What if I miss a few days? Or a few weeks?
Nothing happens. Nothing is lost, nothing resets, no counter goes back to zero.
Come back whenever. The jar will still be full of everything you kept.
Your practice starts today.
Three minutes and a little intention.
Go to my journal →Start tonight.
One line about today is enough. No card, no setup, no streak to protect.
Start journalling free →Rather look first? See a real journal and what Kansha made of it →
Not ready yet?
Free forever · No ads · Private by design