Which company has the best AI model end of September?

Platform comparison

Market context

The market currently prices a 89% probability that a single company will own the model ranking highest on Arena.ai's Text leaderboard by end-September 2026. Settlement hinges on the specific rank column under the "Text Arena | Overall" tab with style control disabled, checked at 12:00 PM ET on 30 September. This is a straightforward ownership attribution question: whichever organisation controls the top-ranked model at that moment determines the outcome. On Polygon, traders holding YES tokens benefit if any model maintains or reaches that position; NO holders profit if the leaderboard shifts to a different company's model or if multiple models tie at rank one.

Historical precedent suggests leaderboard volatility matters more than absolute model capability. OpenAI's GPT-4 and Anthropic's Claude have traded top positions on Arena.ai multiple times since the platform's inception, with leadership shifts often following major model releases rather than incremental improvements. Google DeepMind's Gemini and xAI's Grok have periodically challenged the duopoly. The 89% YES probability reflects confidence that either OpenAI or Anthropic will hold the top spot, though neither company has guaranteed dominance across the nine-month window.

Key catalysts include scheduled model releases from major labs—OpenAI typically announces updates in spring and autumn, whilst Anthropic has maintained a more irregular release cadence. Any surprise release from Meta, Google, or emerging competitors could reshape rankings within weeks. Arena.ai's evaluation methodology itself carries weight; changes to the leaderboard's weighting or judging criteria could alter rankings without new model deployments. Traders should monitor industry announcements closely through Q3 2026, as the September 30 snapshot date leaves minimal reaction time for late-stage surprises.

Methodology

We track Which company has the best AI model end of September? across the five venues with material prediction-market liquidity. The probability shown is the live Polymarket mid; the comparison rows summarise how each venue treats the underlying contract — fees, KYC thresholds, settlement currency, deposit options. The highlighted row marks the cheapest route into Polymarket's order book.

Resolution & payout

At resolution the UMA oracle takes over: a proposer posts the outcome with a bond, any token holder can dispute within two hours. Without dispute the result is accepted and the smart contract distributes USDC instantly.

On Kalshi (CFTC-regulated) resolution runs through their in-house clearing engine in USD. Betfair Exchange settles after match end in the account's local currency. Manifold pays no cash — only its in-platform "mana" currency.

FAQ

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 Polymarket cost to trade?

Polymarket itself charges 0% — the only cost is the Polygon network fee, typically under $0.01 per transaction. Off-chain venues like Kalshi or Betfair charge 2-7% commission.

Do I need to KYC for this market?

On Polymarket directly, no — it's wallet-based. Intermediary brokers like PolyGram trigger KYC only above $1,500 of lifetime trading volume; under that you trade pseudonymously with a single wallet address.

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.