Submit a prompt
Bring a question when the public Arena opens.
PUBLIC ARENA — COMING SOON
Intelligence is only as good as its answers.
Put the world's AI models to the same test. Compare their responses, choose the better answer, and see which intelligence comes out on top.
SOSNET / FIELD STUDY 01A / B
01—02
01—02
Independent comparison, decided by people.Explore the battle archive
THE FORMAT 01 — 02
Good answers should
stand on their own.
ILLUSTRATIVE COMPARISONSTATIC EXAMPLE · NOT MODEL OUTPUT
“What makes a question worth asking?”
DEMO RESPONSE AStart with the kind of curiosity that makes you pause. A worthwhile question opens a door you didn’t know was there.ILLUSTRATIVE SAMPLE
DEMO RESPONSE BOne that changes what you notice next. The answer matters, but the better question gives you a new way to look.ILLUSTRATIVE SAMPLE
DEMO PREVIEW
THE MODEL FIELD
Every answer
leaves a mark.
Rankings begin once live battles and community votes are available. No synthetic results, just the signal earned in the arena.
MODELPROVIDERSTATUSRATING
No completed public results yet.
View the leaderboardA SIMPLE MEASURE
How SOSNET
works.
Compare two responses
Two models will answer the same prompt independently.
Vote for the winner
Choose an answer before model identities are revealed.
Shape the rankings
Recorded votes will shape the public Elo rankings.
A FAIRER COMPARISON
Read the answer.
Then meet the model.
When public battles open, model identities will stay hidden during comparison and appear only after a valid vote.
Explore the demoPUBLIC ARENA — COMING SOON