Who Claims the #3 AI Model Spot by End of May?

The fight for the top spot in AI gets most of the headlines. The battle for third place is where the real drama lives right now. Anthropic sits at 51.5% implied probability on Polymarket, meaning the market sees this as a genuine coin-flip with a slight lean toward Claude holding the bronze position by May 31. That razor-thin edge reflects how compressed the AI model rankings have become across the top four or five players.

The contract asks a specific question: which company holds the third-ranked AI model by end of May 2026, scored on style-controlled benchmarks. Anthropic’s YES contract trades at $0.52. The field of competitors includes Google, Meta, xAI, OpenAI, DeepSeek, Mistral, and several Chinese labs. With $2,021 in total volume and a 24-hour spike of $1,383, almost all trading activity hit in the last day. That pattern usually means a specific catalyst moved someone to act.

How the Anthropic Contract Works

The YES contract pays out if Anthropic’s top model ranks third among all AI systems on the style-controlled benchmark used for resolution on May 31, 2026. The resolution source is market resolution, which means judges or an established leaderboard snapshot determines the final ranking.

• YES ($0.52, 51.5% implied): Anthropic’s Claude model ranks exactly third on the style-controlled AI benchmark at end of May.
• NO ($0.49, 48.5% implied): Any other company’s model occupies third place, or Anthropic ranks higher or lower than third.

A NO outcome resolves in favor of the field if Google Gemini, Meta Llama, xAI Grok, OpenAI GPT, or a Chinese lab model pushes Claude out of the third slot entirely. Anthropic finishing second would also trigger NO. The contract is binary and ranking-specific, so a strong Claude release that vaults to second place pays nothing on this contract.

Sponsored Partner

Market Signals and What the Numbers Say

The momentum composite reads cautious optimism. The 1-hour change is flat at +0.0%, the 24-hour change is +2.0%, and the trend score sits at 26.54. That combination points to mild buying pressure with no acceleration. The 24-hour volume of $1,383 against a total market volume of $2,021 means nearly the entire trading history of this contract happened in one day. Something triggered that burst, likely chatter around a benchmark update or a model release announcement from one of the competing labs.

Liquidity stands at $8,951, which is thin. For a contract expiring May 31, that means any meaningful position size will move the price. The total volume of $2,021 puts this firmly in low-conviction territory. Traders are watching this one, not loading up on it.

Key factors shaping the contract:

• Anthropic’s Claude 3.7 Sonnet currently scores near the top of Chatbot Arena style-controlled leaderboards, which directly informs how this contract likely resolves.
• Google’s Gemini 2.5 Pro has been trading blows with Claude for the second and third positions on multiple benchmarks as of late April 2026.
• Meta’s Llama 4 Scout and Maverick releases in April 2026 added a credible challenger for the third slot, particularly on open-weight evaluations.
• The 1-hour price change of +0.0% and 24-hour change of +2.0% together indicate deceleration after the initial buy-in, not a momentum run.
• xAI’s Grok 3 has been competitive on reasoning tasks but has not consistently cracked the top three on style-controlled benchmarks.

Anthropic’s Position and the Risks Around It

Anthropic has earned its 51.5% by running Claude 3.7 Sonnet into a strong stretch of benchmark performance. On Chatbot Arena’s style-controlled leaderboard, Claude models have held top-three positions for several months. The key advantage for Anthropic is consistency. Claude does not swing wildly between benchmark snapshots, and that stability matters when a single end-of-May reading determines everything.

The risk is Google. Gemini 2.5 Pro has been closing the gap on style-controlled tasks specifically, which is the exact evaluation type this contract uses. Google has the infrastructure and the model iteration speed to push a new version before May 31. If Gemini 2.5 Pro or a follow-up release lands a benchmark score that slots above Claude, the third-place ranking shifts instantly. Meta’s Llama 4 family is a secondary threat, especially if a larger Maverick variant posts strong style scores before the resolution date.

Signals to monitor before May 31:

• Any Chatbot Arena leaderboard update showing Gemini 2.5 Pro crossing Claude on style-controlled categories would reprice this contract sharply toward NO.
• Anthropic releasing a Claude 4 preview or an updated Sonnet variant before the deadline would push YES higher immediately.
• Meta publishing Llama 4 Maverick full benchmark results with style-controlled scores above Claude would pull probability toward NO.
• xAI announcing a Grok 3.5 or similar release with a competitive style benchmark would introduce another variable into the third-place math.
• Any leaderboard methodology change at Chatbot Arena or the resolution source would create ambiguity and likely compress both YES and NO prices toward 50 cents.

The $2,021 total volume is the honest signal here. Conviction is low on both sides. The market has landed at near-even odds because the evidence genuinely supports a near-even split. Anthropic’s track record earns a slight edge, but the window to May 31 is long enough for Google or Meta to flip the ranking.

LINES VERDICT

Anthropic Slight Favorite, Field Still Live

Claude’s consistency on style-controlled benchmarks gives Anthropic a real but fragile edge. One Google or Meta model update before May 31 is enough to close this gap entirely.

What the market says: Anthropic trades at 51.5%, barely above a coin-flip, reflecting genuine uncertainty about a ranking that could shift with any major model release before the May 31, 2026 resolution date.

FAQ

What does 51.5% mean for this contract? The market puts roughly even odds on Anthropic holding third place. A $1 YES contract returns about $1.92 if Anthropic finishes exactly third on the style-controlled benchmark at resolution.

What does the NO contract cover? A NO contract pays out if any company other than Anthropic holds the third-ranked AI model spot on May 31. That includes Google, Meta, xAI, OpenAI, or any other lab in the field.

What moves the price on this contract? Benchmark leaderboard updates, new model releases from Google or Meta, and Anthropic announcements about Claude updates are the primary catalysts. Any of those events can shift the price several percentage points overnight.

When and how does this market resolve? The contract resolves on May 31, 2026, based on the style-controlled AI model rankings at that date. The specific leaderboard or panel used for resolution determines the final outcome.

Is the volume reliable enough to trust the price? Volume is $2,021 total with $8,951 in liquidity. That is a thin market. The price is directionally informative but can move significantly on small trades, so treat the 51.5% figure as a rough signal rather than a precise probability.