Latest Kalshi Market Pricing

To review the latest market pricing, visit Kalshi:

Market Context

QUICK PICK

Bet YES on OpenAI holding a top-ranked AI model before 2027 at 63.5 cents. The $200 million Pentagon contract, Amazon's $50 billion pivot toward OpenAI, and Anthropic's sweeping federal ban make OpenAI the structural favorite in this race.

THE SETUP

This market asks whether the fallout from Anthropic's Pentagon standoff reshapes the AI competitive landscape before 2027. The Pentagon declared Anthropic a supply-chain risk to national security after the company refused to sign military terms by the February 27, 2026 deadline. That designation bars any contractor, supplier, or partner that does business with the U.S. military from conducting commercial activity with Anthropic. Within hours, OpenAI CEO Sam Altman announced an agreement to deploy models in the Department of War's classified network, building on $200 million in prior unclassified contracts.

THE PAYOFF

Buy OpenAI at 63.5 cents with a clear catalyst already in place. Any further defense contract expansions or Amazon-OpenAI integration announcements would push the 63.5 cent price higher fast.

Bonus: Watch the IPO market closely: OpenAI leads the Anthropic race at a 54-cent midpoint on Kalshi, while Anthropic's top YES bid sits at 52 cents, keeping the gap narrow but real.

Kalshi Contract Analysis: LMSYS Chatbot Arena #1 LLM Ranking

1. What You're Betting On

You're betting on whether a specific AI company (designated by Kalshi) will have the #1 ranked large language model on the LMSYS Chatbot Arena's Overall Arena Elo leaderboard (UB category) at any point during a specified time window. If their model hits the top spot solo — even for a single day — the contract pays out Yes.

2. How It Resolves

Yes ($1.00): The designated company has an LLM ranked #1 on the Overall Arena Elo (UB) leaderboard at 10:00 AM ET on any day between the contract's issuance and its end date.

No ($0.00): The company's LLM never reaches the sole #1 spot during that window, or the end date arrives without it happening.

What counts: The model must be ranked strictly #1 — top of the leaderboard, no ties. The check happens at 10:00 AM ET each day against whatever the leaderboard shows at that moment.

What doesn't count: A tie for #1 with another model does not satisfy the payout criterion. The company's model must be the undisputed top-ranked LLM. Also, any revisions to the leaderboard made after the contract expires are ignored — the snapshot at expiration time is final.

Who decides: LMSYS is the Source Agency. The specific leaderboard used is the Overall Arena Elo (UB) rankings on the Chatbot Arena Leaderboard.

3. Key Dates & Times

  • Monitoring period: Every day from issuance until the end date (specified by Kalshi per contract iteration), checked at 10:00 AM ET.

  • Market closes (last trading): The sooner of (a) 10:00 AM ET on the first day the payout criterion is met, or (b) 11:59 PM ET the day before the end date.

  • Expiration date: The sooner of the first 10:00 AM ET after the #1 ranking occurs, or the contract's end date.

  • Expiration time: 10:00 AM ET.

  • Settlement: No later than the day after expiration, unless under review.

Important timing nuance: This contract can resolve early. If the designated company's model hits #1 on a Tuesday and the leaderboard reflects that at 10:00 AM ET Wednesday, the contract expires and trading stops at that 10:00 AM ET Wednesday. You don't have to wait for the end date.

4. Quirks & Edge Cases

  • Ties kill a Yes. This is the single biggest gotcha. If the company's model is tied at #1 with any other model, that does not count. It must be the sole #1. On a leaderboard where Elo scores can be very close, this matters a lot.

  • "UB" designation matters. The contract specifies the "Overall Arena Elo rankings (UB)" — make sure you're looking at the correct leaderboard variant on the Chatbot Arena site. LMSYS has multiple ranking categories and methodologies; only UB counts here.

  • First publication only. If LMSYS later revises their rankings (recalculates Elo scores, removes contaminated votes, etc.) after the contract has expired, those revisions are irrelevant. Whatever the leaderboard showed at 10:00 AM ET on the expiration date is the final answer.

  • Daily snapshot, not continuous. The check is at 10:00 AM ET. If a model briefly hits #1 at 3:00 PM but drops by the next morning's 10:00 AM check, it may not count — the contract looks at the leaderboard state at the specific check time.

  • Placeholders for company and end date. The contract uses placeholders ("a group specified by the Exchange" and a calendar date specified by Kalshi). When you're trading a specific iteration, Kalshi fills these in. Make sure you know which company and which end date your specific contract refers to.

  • Early resolution is built in. The moment the criterion is met at a 10:00 AM ET check, the contract expires and settles. This means Yes holders get paid quickly, but it also means trading stops immediately.

  • Position limit: $25,000 worth of contracts per strike per member.

  • Kalshi review: Kalshi reserves the right to review the outcome before settlement under Rule 6.3(c).

5. Who Can't Trade This

  • Employees of LMSYS

  • Anyone with material non-public information about the leaderboard rankings (e.g., someone who knows about upcoming ranking recalculations before they're published)

  • Anyone who can directly influence the outcome (e.g., someone who controls which models appear on the leaderboard or how Elo scores are computed)

6. Bottom Line for Traders

Watch the LMSYS Chatbot Arena Overall Arena Elo (UB) leaderboard daily — that's your single source of truth. The most critical thing to understand is the no-ties rule: the designated company's model must be the undisputed #1, not merely tied for first. If you're trading this, keep an eye on the Elo gaps between the top models, because a few-point difference on the leaderboard is the difference between a payout and nothing. And remember, the contract can snap shut early the morning after the criterion is met, so don't assume you'll have until the end date to exit your position.