Which company has best AI model end of June?

Platform comparison

Active sub-markets

Market context

The Chatbot Arena Leaderboard ranks large language models through head-to-head user voting, with the top-ranked model's owner determining the market resolution on 30 June 2026. Current standings show OpenAI's GPT-4o and Anthropic's Claude 3.5 Sonnet competing for the top position, though the leaderboard shifts regularly as new model versions release and voting patterns evolve. The resolution mechanism depends entirely on the "Rank" column with Arena score as tiebreaker, making incremental leaderboard movements consequential for the outcome.

Historical precedent suggests dominance concentration among three firms. Over the past eighteen months, OpenAI, Anthropic, and Google have occupied the top three positions with minimal disruption from other developers. When new model releases occur—particularly major version updates—they typically enter the leaderboard within weeks and can rapidly accumulate voting data. The 14% implied probability reflects uncertainty about which organisation will maintain or achieve top rank by mid-2026, rather than scepticism about whether a clear leader will exist.

Traders should monitor scheduled model releases from all three contenders through early 2026. OpenAI's product roadmap, Anthropic's deployment announcements, and Google's DeepMind initiatives all carry weight. The leaderboard's sensitivity to recent voting means models released closer to June will have fresher user preference data, potentially advantaging late-cycle launches. Additionally, changes to Arena's voting methodology or user base composition could shift rankings independent of underlying model capability, though such alterations remain infrequent.

Methodology

Methodologically we separate two layers: the live probability (Polymarket mid-price) and the platform attributes (fee, KYC, settlement currency, payment rails). The odds column is filled only where we have clean data — that avoids the made-up numbers that get a network demoted when search engines cross-check against the source venue.

Resolution & payout

Polymarket-based markets settle through the UMA Optimistic Oracle on Polygon. A proposer submits the outcome, a two-hour challenge window opens, and unchallenged proposals finalise the resolution. Payouts settle automatically in USDC the moment the result is final — no bookmaker, no delay.

Kalshi-based markets settle in USD via the CFTC-regulated clearinghouse. Betfair Exchange settles in GBP/EUR net of commission. Manifold is play-money and does not pay out real funds.

FAQ

Is this market available outside the US?

Sport Prediction is available in most jurisdictions where Polymarket isn't directly accessible. Polymarket itself is geo-blocked in the US/UK/EU. Always check local regulations.

How does resolution work?

Through the UMA Optimistic Oracle on Polygon: a proposer submits the outcome, a two-hour challenge window opens, and USDC payouts settle automatically once the result is final.

What's the difference between YES and NO shares?

A YES share pays $1.00 if the event happens, $0 otherwise. A NO share pays $1.00 if the event doesn't happen. The market price between 0¢ and 100¢ is the implied probability.

What does it cost to trade on Sport Prediction?

Zero. Sport Prediction routes every order to the live Polymarket order book; the only cost is the Polygon network fee, typically under $0.01 per transaction.

How reliable are the quoted odds?

The YES/NO percentages are the live mid-prices of the Polymarket order book. On deep markets they move every few seconds; on thinner ones you'll see short plateaus.