Settle any argument.
Ask anything, 4 models fight for the best answer, you get the verdict. Our advanced debate system generates a more reliable answer than a single chatbot.
🎁 New here? Your first debate is FREE — a Quickie or a photo-judging Vision. Your pick.
Example Vision verdict
“Rate my sourdough out of 10 — be honest”
The verdict
7/10 — proper crust, honest crumb, needs a bolder score.
crust8
crumb6
shape7
ambition9
7/10
Three models argued it out: GPT called the scoring “generous for a first loaf”, Grok defended the open crumb, and the judge landed on 7 — not an average, a negotiated ruling.
Panel confidence64%
Vision test rig
Internal only. Picks a photo, resizes to 2048px, runs a 50-credit Vision debate, shows the raw result.
How it works
One chatbot is one opinion — and chatbots can confidently get things wrong. Mass Debates makes four AI models argue it out: each answers blind (no copying the confident one), then they attack each other’s reasoning, and a judge hands down one verdict with an honest confidence score. Answers are cross-checked, not just generated.
Quick verdict — straight blind answers, quick ruling. The daft-argument settler.
Full debate — adds the clash round: rebuttals, concessions, and devil’s advocates stress-testing the consensus.
Power — the full clash plus a fact-check pass rating the claims the answer actually rests on.
🎁 First debate free — every new member’s first Quickie or Vision is on the house.
Vision — upload a photo and the panel judges it out of 10: rate your date, your cake, your work. Scorecard included.
✉ Contact us
Sign in
We'll email you a 6-digit code — no password to remember.
✓ No password ✓ No forms ✓ Email + code, done
Add credits
1 credit = 1p. Pick an amount, then pay in a couple of taps — no leaving the page.
1 credit = 1p · a Quickie is 5, a Vision verdict 30.
The motion
🎲 Random question
Full debate
Blind answers, two seats argue the other side, a clash round, then a verdict.
Active: