←All eventsmonthly series · prediction markets

#3 AI model Style Control On: past results & live odds

#3 AI model Style Control On is a monthly series of prediction markets listed on Polymarket. This page tracks the event trading now, the results of the last 10 that resolved — with the odds the market gave the winner 24 hours before close — and how often the favorite of that moment actually won. Over the last 10 resolved events where the price history proves who the favorite was 24 hours before close, the favorite won 10 of 10 (100%), priced at 95% on average.

Trading now

#3 AI Lab end of October? (Style Control On)

Anthropic 49% · Meta 33% · Mistral 6% — resolves Oct 31, 2026 · $68K traded

Recent results

EventResolvedWinnerWinner's odds 24h beforeFavorite 24h before
#3 AI Lab end of September? (Style Control On)Sep 30, 2026Google96%Google 96% — won
#3 AI Lab end of August? (Style Control On)Sep 28, 2026Google93%Google 93% — won
Which company has the #3 AI model end of July? (Style Control On)Jul 31, 2026Anthropic>99%Anthropic >99% — won
Which company has the #3 AI model end of June? (Style Control On)Jun 30, 2026Anthropic98%Anthropic 98% — won
Which company has the #3 AI model end of May? (Style Control On)Jun 1, 2026Anthropic99%Anthropic 99% — won
Which company has the #3 AI model end of April? (Style Control On)Apr 30, 2026Anthropic93%Anthropic 93% — won
Which company has the #3 AI model end of March? (Style Control On)Mar 31, 2026Google94%Google 94% — won
Which company has the #3 AI model end of February? (Style Control On)Feb 28, 2026Google90%Google 90% — won
Which company has the #3 AI model end of January? (Style Control On)Jan 31, 2026Google97%Google 97% — won
Which company has the #3 AI model end of December? (Style Control On)Dec 31, 2025xAI86%xAI 86% — won

Sources: settlements and events from Polymarket's public Gamma API; odds 24 hours before close (the earlier of the scheduled end and the actual close) from Polymarket's public CLOB price history. A “—” means the price history doesn't prove it, so that event is left out of the favorite count rather than estimated.