This rizz test scores what you actually text. Paste an opener or a whole chat and AuraMog checks your side for five things: detail, hook, vibe, originality and flow. Each part is worth up to 20 points and the parts add up to your score out of 100. Run the same text twice and you get the same number.
How does the rizz test score your texts?
The model reads your text and answers fixed questions about it: does it name something specific, does it give them an easy answer, is there any play in it, how many stock phrases it leans on. It never picks the number. A fixed formula works like a rizz calculator and turns those answers into points, so the same text lands on the same score.
| Part | What earns points | What costs points |
|---|---|---|
| Detail | Something only they have: a photo, a prompt answer, a thing they said | Lines that could go to anyone |
| Hook | One easy, fun thing to answer | Nothing to answer, or a job interview question |
| Vibe | Play, teasing, a light challenge | Dry replies, all caps, a pile of emoji |
| Originality | Words that sound like you | Stock openers, pickup lines, leading with looks |
| Flow | A length they can read at a glance, a pace that matches theirs | One-word replies, walls of text, three messages in a row |
What do the rizz levels mean?
- W rizz, 90 and up. Specific, playful and easy to answer.
- Solid rizz, 75 to 89. It works. One part is holding it back, usually detail or hook.
- Mid rizz, 55 to 74. Polite and easy to forget. It gets an answer, usually small talk.
- Low rizz, 35 to 54. The other person has to do all the work to keep it going.
- L rizz, 1 to 34. Home of hey, wyd and copied pickup lines.
Do you have rizz? What high-scoring texts have in common
How much rizz do you have? A high score comes from habits you can check in your own texts before you send them.
- They name one real detail about the other person, taken from a photo, a prompt or something said earlier.
- They end on one light question or a choice between two things.
- They tease a choice or a take and leave looks out of it.
- They fit on a phone screen without scrolling.
- In a chat, they answer the question that was asked before asking a new one.
What drags a rizz score down?
- Opening with hey, hi or wyd. It is the most common opener on every app and it scores near the bottom.
- A compliment on looks as the whole message.
- Interview questions one after another: what do you do, where are you from, what are you looking for.
- Double and triple texting before they answer.
- Leaving their question unanswered.
- Pressure of any kind. Guilt, ultimatums and demands to reply get no score at all.
Rizz test examples: five real lines and their scores
Each line below went through the same test you see above. Next to it are its score, the points for each part and, when one part drags it down, the note for that part.
- hey 13/100, L rizz. Detail 2, hook 2, vibe 5, originality 0, flow 4. Originality is the weak spot. The most common opener on every app.
- hey beautiful, how's your day going? 52/100, Low rizz. Detail 9, hook 13, vibe 12, originality 0, flow 18. Originality is the weak spot. Stock phrases they've seen a hundred times today.
- How was your weekend? 50/100, Low rizz. Detail 9, hook 13, vibe 5, originality 13, flow 10. Vibe is the weak spot. Reads dry. Nothing to smile at.
- What's the best thing you've eaten this month? 79/100, Solid rizz. Detail 9, hook 20, vibe 12, originality 20, flow 18. Detail is the weak spot. There's a topic, but it fits any match.
- The dog in your third photo looks like he runs the house. Does he approve of strangers? 96/100, W rizz. Detail 20, hook 20, vibe 18, originality 20, flow 18.
Rizz quiz or rizz test: which one tells you more?
A rizz quiz asks how you think you flirt and hands back a label. This test reads the message itself, so the result points at the exact words to change. Take a quiz for fun, then run your last opener here and compare.
The two numbers answer different questions. The quiz measures how you see yourself. The test measures what the other person actually reads.
How do you raise your rizz score?
- Fix the lowest part first. The result shows which of the five parts cost you most.
- Swap the greeting for a detail. Start with the thing you noticed and drop the hey.
- End on one question. Cut every question after the first.
- Read it out loud. If it sounds like a pickup line, rewrite it in your own words.
- In a chat, match their length and let them answer before you send the next one.
Stuck on the rewrite? Rizz translator turns a flat line into a flirty one in your own voice.
Frequently asked questions
It is consistent, and that is the part you can check. The model answers fixed questions about your text and a formula turns the answers into points, so the same text gets the same score. It measures how easy and fun your message is to answer. It cannot tell you how one specific person will react.
The model only describes the text and a fixed formula picks the number. Once a text is scored, the result is kept under a fingerprint of that exact text, so pasting it again returns the same number and the same roast without asking the model twice. Change one word and it counts as a new text.
Yes. Switch to Whole chat and paste it, or add up to three screenshots. Only your messages are scored: detail, questions, playfulness, stock phrases and whether you match their pace. Their messages are read for context and never quoted in the roast or on the share card.
It gets no score. Guilt trips, ultimatums, demands to reply and insults get a plain note on how they land instead of a number, and there is no share card for them. Rewrite the line so the other person can answer or say no easily, then run it again and see where it lands.
No. Each of the five parts is worth up to 20 points, but the formula never gives all five full marks at once, so no text reaches 100. Even a great opener has a version that fits the person better. W rizz starts at 90, and a score there is worth sharing.
A chatbot rater or a rizz meter writes a new number every time you ask, so the same line can come back 75 and then 72. Here the model never writes the number. It answers fixed questions about your text, and the score comes from the formula described on this page, one part at a time.
