Take it together on one phone. You and your partner answer 11 questions — about seven minutes — and Overchat AI compares your two answers side by side to build a single couple report: a compatibility score, a breakdown across nine metrics, your couple archetype, and a green-flag/red-flag count for each of you. The score comes from what the two of you actually answered, not from a birth chart or a name calculator.
Most compatibility tests online score one person guessing about the relationship, or match two birth dates. This one needs both of you in the room. Every question is answered twice — once by her, once by him — and the report is built from the distance between the two answers.
Pass the phone back and forth through 11 questions. Each answer is tagged to the partner who gave it, so the compatibility score measures the gap between your two versions of the relationship instead of averaging one person's opinion.
Your compatibility percentage comes with nine separate metrics: trust level, values match, humor match, vibe match, sexual compatibility, freedom to be yourself, stored-up irritation, toxicity level and breakup risk. You see exactly which one dragged the number down.
Nothing to install and no account needed to start. Answer the last question and the couple report loads on the same phone you took the test on.
The two of you answer, and the report does the rest: a straight answer to “are we compatible?”, a score on nine metrics, a flag count for each partner, a forecast, and a fun half you will read out loud. Here is what lands on screen when the relationship quiz ends.
The report opens with one percentage for the two of you, then breaks it into nine scores you can actually argue about: trust level, physical chemistry, values match, humor match, toxicity level, stored-up irritation, freedom to be yourself, overall vibe and breakup risk. The compatibility quiz is scored the same way every time, so the number still means something when you retake it in six months.
The SHE vs HE battle settles the arguments the two of you have been having for years: who loves harder, who is actually in charge, who understands the other better, who is the couple's buzzkill, who is the chief manipulator — and who is more toxic in a fight. Next to it, the flag-o-meter counts green and red flags for each of you separately, built from how each partner described the other.
The report names the single thing most likely to end the two of you, explains why in plain language, and puts numbers on it: probability of a big fight, probability of cheating, breakup risk. Then it forecasts how much longer the relationship lasts on its current trajectory. Uncomfortable to read together, which is the point — it names the one thing to fix while fixing it is still cheap.
Which movie your relationship is. Which cartoon duo the two of you are. Which cat meme describes your couple. Then seven stress tests — a newborn in the house, money trouble, long distance, moving abroad — each with a verdict from dream team to danger zone, and a date idea engineered around whichever metric you scored worst on.
USE CASES
One percentage for the two of you, calculated from how far apart your answers landed across all 11 questions. The scoring is fixed, so you can retake the compatibility test in six months and see whether the number actually moved.
The report names the kind of couple you are, then points at your risk zone — the area most likely to start your next fight — and your relationship superpower, the thing the two of you handle better than most couples.
The flag-o-meter counts flags for each partner separately, based on how each of you describes the other during a fight. Talking it out or taking a pause counts green; guilt-tripping, cold silence and making you doubt your own memory count red.
A direct comparison built from the answers you both gave: who loves harder, who is really in charge, and who is the chief manipulator in this relationship.
Your couple is run through seven scenarios — a newborn in the house, money trouble, long distance, moving abroad and three more — with a verdict on each, from dream team to danger zone.
The weak spots come with something to do about them, including a date idea built around the exact metric your couple scored lowest on.
Grab your partner and one phone. The compatibility test takes about seven minutes and your score is free.
The score is a weighted comparison of two sets of answers, not a mood reading. Four things feed it, and moving closer on any one of them moves the number. The score itself is free; the full couple report is part of Overchat Pro.
| Criterion | What it measures | What the score uses |
|---|---|---|
| ✓Values match | Whether the two of you pick the same side when the choice actually costs something | Would-you-rather round, compared answer by answer — the heaviest single input |
| ✓Boundaries | Where each of you draws the line on the same situation | FINE-or-WEIRD blitz, item by item + who called a risky one fine |
| ✓Shared traits | How far apart your two descriptions of this couple are | Overlap between the traits she picked and the ones he picked |
| ✓Flags & tone | How each of you behaves in a fight, and how you write about each other | Green-to-red flag ratio, plus warm and negative language in the written answers |
Overchat AI brings you the power of the world's top AI models: ChatGPT, Claude, Gemini, Mistral, and more.
You and your partner enter your names and birth dates, then answer 11 questions on the same phone — open questions about the relationship, multiple-choice questions about how the other one behaves, a would-you-rather round and a rapid-fire FINE-or-WEIRD blitz. Overchat AI compares each pair of answers and turns the gaps between them into a compatibility score with a full couple report.
Compatibility has less to do with sharing hobbies and more with how the two of you handle friction: whether an argument ends in repair, whether you can be yourself around each other, whether your values line up on the things nobody wants to compromise on. That is what this test measures — trust, values, humor, vibe, intimacy, freedom, stored-up irritation and toxicity — instead of asking whether you both like the same movies.
The useful ones are the ones both partners actually take. A quiz filled in by one person records how that person sees the relationship, which makes the result a self-assessment rather than a compatibility score. This test is built for two people on one phone: every question is answered twice, and the report is built from the places where the two answers disagree.
The score is calculated, not random — the same answers always produce the same number, so a retake is a real comparison rather than a new roll of the dice. What it measures is how closely your two answers line up right now across nine relationship metrics. Read it as a starting point for a conversation about the low-scoring areas, not as a verdict on the relationship.
The 7-7-7 rule is a scheduling habit for long-term couples: a date night every 7 days, a night away together every 7 weeks, and a longer trip every 7 months. The point is to book time together before the calendar fills up with everything else. If your report comes back with a low vibe match or a lot of stored-up irritation, this is the kind of routine that moves those two numbers.
The 3-6-9 rule marks three checkpoints in a new relationship. Around 3 months the early infatuation fades and real habits start showing. Around 6 months you have seen how the other person handles stress and conflict. Around 9 months you have enough evidence to decide whether this is long-term. A compatibility test tells you most after the 3-month mark, once there is something real to answer about.
Your answers are used to build your couple report and nothing else. Nothing is published, and you decide whether to show the result to anyone.
The 2-2-2 rule is a lighter version of the same idea: a date every 2 weeks, a weekend away every 2 months, and a week-long holiday every 2 years. Couples use it to keep time together on the calendar instead of hoping it happens on its own. If the report puts your risk zone in routine or attention, this is a cheap place to start.
Chinese company Baidu has quietly released a model that's just as good as DeepSeek and ChatGPT — ERNIE 4.5 delivers exceptional performance at a fraction of the cost — and most people outside China have never even heard of it.
If you're wondering: what is DeepSeek AI, the answer is quite simple. DeepSeek AI is a Chinese AI research lab that builds large language models. The company released several models in late 2024 and early 2025 that compete directly with OpenAI's GPT-4 and Anthropic's Claude, but at a fraction of the operating cost.
ChatGPT Plus costs $20 per month plus tax, while the Pro plan costs $200 per month plus tax. Understandably, not everyone is willing to pay that much for a chatbot, especially if the premium features aren’t worth it to them. This raises an obvious question: how can I use ChatGPT for free?
Available on Web, iOS, and Android. Access your AI assistant anywhere, anytime.
.svg){kind=icon}
