The temperature
A 0–100 read of how heated it got, from "Cool-headed" to "Boiling over" — useful for noticing whether you two escalate or simmer.
Every argument is two conversations: the one you had, and the one you were actually having. Paste any chat — a WhatsApp export, an iMessage thread, whatever you've got — and get a warm, therapist-style read of both. No winner declared. Nobody called toxic.
An AI reflection, not therapy or a verdict. Takes about a minute.
Paste it straight from your chat app, or upload the .txt export. Names and all — we'll use them kindly.
That was a long one — we kept the last part (the most recent 15,000 characters), which is usually where the argument lives.
Analyzed in the moment. Never stored. Never used for training.
Reading the room…
This usually takes ten to twenty seconds.
The full debrief is below — what it was really about, the patterns underneath, and one gentle way to reopen it. Want a copy for later, too?
Two people fought because two people care. Here's what we heard.
An AI-generated reflection, not therapy, a diagnosis, or a verdict. If things feel heavier than a rough patch, please reach out to a licensed professional.
Patterns, not accusations. Every couple has them; naming them is how they loosen.
Small reaches back toward each other — easy to miss mid-fight, worth celebrating after.
One sentence either of you could send to reopen this gently.
This debrief happened after the fight. Twogle is there during — a private space for each of you, a shared one for both, and gentle prompts that catch the pattern while it's still just a conversation.
A debrief reads one argument. The Conflict Style Quiz reads how the two of you fight in general — and what happens when your style meets your partner's.
Take the Conflict Style Quiz →A free tool that reads a couple's argument the way a good therapist would: warmly, on both sides, and with an eye for what the fight was actually about. You paste the chat — or upload a WhatsApp .txt export — and get back a debrief: the emotional temperature, the surface topic versus the real issue underneath, each person's communication style and unmet need, the patterns that kept the fight going, the repair attempts you both missed, and one sentence either of you could send to reopen the conversation gently. It never picks a winner, never diagnoses anyone, and never calls either of you toxic. Most arguments are two people making a reasonable point badly; this is the translation.
A 0–100 read of how heated it got, from "Cool-headed" to "Boiling over" — useful for noticing whether you two escalate or simmer.
What the fight looked like it was about, and what it was really about. These are almost never the same thing, and the gap is where the fix lives.
Each person's communication style in this argument, the need that went unmet, and — because it is always there — one thing each of you did well.
The moves that kept the fight going (the "always", the retreat, the scorekeeping), each with a short quote from your chat and a gentler way to say the same thing.
The moments one of you reached back toward the other — a softening, a joke, an olive branch. Couples who notice these fight the same amount but recover faster.
Three concrete suggestions and one sentence either partner could send to reopen the topic without reopening the wound.
Here is the kind of argument people paste, and the kind of debrief they get back. (Hand-written example — your debrief is generated fresh from your own conversation.)
Maya: you said you'd book the flights this week
Sam: I've had a brutal week, I'll get to it
Maya: you always say that. I end up doing everything for this trip
Sam: "always". great. here we go
Maya: I'm not starting anything, I just don't want it hanging over us
Sam: fine. I'll do it tonight. happy?
Maya: honestly not really. it's not about the flights
58 — Heated. It warmed up fast around "always", then cooled slightly when Maya softened. Nobody boiled over.
Booking flights for the trip.
Maya wanting to feel like Sam shares the mental load of the trip, and Sam wanting his hard week to be seen before being handed a task.
"Always" and "never". "You always say that" turned one late booking into a character verdict, which is the moment Sam stopped hearing the request. Gentler version: "This is the third time it's slipped, and it's making me feel like the trip is mine to carry."
"I'm not starting anything, I just don't want it hanging over us" — Maya reaching for de-escalation. It sailed past Sam, which is normal; most repair attempts do.
"Can we try the flights conversation again? I don't think either of us said the real thing."
You are pasting a real fight with a real person you love, so we built this the only acceptable way: your conversation is analyzed in the moment, never stored, and never used for training. There is no account, no database row, no log line containing your words, and no copy kept after the debrief is written. If you email yourself the result, the email carries the summary numbers — never the transcript. The one thing we cannot control is your own clipboard, so paste from a device you trust.
Yes, and we mean it in the strictest sense. Your transcript is analyzed in the moment and never stored: no database, no logs, no training data, no copy kept anywhere. It exists only long enough to write your debrief. Once the debrief is on your screen, the conversation is gone from ours.
Any chat between the two of you, from any app. Paste it straight in, or upload a .txt export (the WhatsApp "Export chat" file works as-is — we strip the timestamps and system lines for you). It needs to be between 80 and 15,000 characters; if it is longer, we keep the most recent part, which is usually where the argument lives.
No, and we want to be plain about that. This is an AI-generated reflection, built on ideas from couples research but not a clinical assessment, a diagnosis, or a substitute for working with a professional. If what is happening between you feels heavier than a rough patch, a licensed couples therapist is the right next step.
No, on purpose. The analyzer never picks a winner, never diagnoses anyone, and never calls either of you toxic. It reads both sides with the same warmth, because most arguments are two people making a reasonable point badly. What you get is a translation, not a verdict.
About a minute. Paste the chat, add a line of context if you want, and the debrief usually takes ten to twenty seconds to write. It is free and there is no signup — we offer to email you a copy at the end, and that is always skippable.