Chatbot Arena Crown by June 30: Anthropic and Google Face Off in LLM Rankings

What is happening now

Polymarket’s grouped event “Which company has the best AI model end of June?” asks which firm will own the top‑ranked LLM on the Chatbot Arena LLM Leaderboard when it is checked on 30 June 2026 at 12:00 PM ET. The market aggregates 15 individual binary contracts, each asking whether a specific company’s model will be #1. As of 2 June 2026 the overall market shows a 99.9 % probability that none of the listed firms will be the leader, reflected by the “No” outcome on the primary market (ID 631139). The only contract with a sizable opposite view is the Anthropic market (ID 631140), where “Yes” trades at ~84 % probability, indicating the market’s consensus that Anthropic’s model is the most likely to finish first.

How the market is structured

This is a grouped multi‑market** event. Each sub‑market is a binary “Yes/No” contract tied to a single company. The primary market (Google) is used for the headline probability; its “No” side (86.5 % price) reflects the chance that Google is not the leader. All other companies have their own contracts, most priced near 0.0015 (≈0.2 % chance) for “Yes”. The Anthropic contract is an outlier, trading at 0.8395 for “Yes”. The market resolves by checking the leaderboard rank at the specified timestamp; the company whose model holds rank 1 wins, and all other contracts resolve “No”.

Leading outcomes

• Anthropic – “Yes” at 0.8395 (≈84 % implied probability).
• All other firms – “Yes” priced between 0.0015 and 0.0035 (≈0.2‑0.4 % chance).
• Primary “No” (Google not leading) at 0.865 (86.5 %); the aggregate “No” across the group is ~99.9 %.

Path to the leading outcome

For Anthropic to win, its model (currently Claude 3‑Opus) must retain or improve its performance relative to rivals through the next six months. Key milestones that would reinforce this view include:

• Release of a new Claude iteration (e.g., Claude 3‑Turbo‑Pro) before Q2 2026 with benchmark scores surpassing OpenAI’s GPT‑4‑Turbo and Google’s Gemini 1.5.
• Public benchmark results on Chatbot Arena showing Anthropic’s model holding the top‑rank for at least three consecutive weeks in May‑June 2026.
• Strategic partnerships (e.g., with Microsoft or Amazon) that expand Anthropic’s deployment, increasing real‑world usage data that feeds the Arena scoring algorithm.

What could change the pricing

Because the market is anchored to a single leaderboard snapshot, any development that shifts the ranking order before 30 June 2026 can swing prices dramatically:

• OpenAI breakthroughs: A major model upgrade (e.g., GPT‑5) released before June could push OpenAI’s rank to #1, collapsing Anthropic’s “Yes” price and inflating the OpenAI “Yes” contract (currently 2.5 %).
• Google’s Gemini release: If Google unveils Gemini 2.0 with a decisive lead in the Arena, the Google “Yes” price (13.5 %) could surge, while the primary “No” price would fall.
• Emerging competitors: Unexpected performance jumps from smaller players such as xAI, Moonshot, or Z.ai (all currently at ~0.2 % “Yes”) would cause their contracts to spike and dilute Anthropic’s dominance.
• Leaderboard downtime: The resolution rule states the market stays open until the leaderboard is back online. A prolonged outage could delay settlement and increase uncertainty, widening spreads across all contracts.
• Scoring methodology changes: If Chatbot Arena alters its evaluation criteria (e.g., weighting safety or latency differently), the ranking could reshuffle, affecting all odds.

Editorial read

The Polymarket event is essentially a bet on which AI firm will top the Chatbot Arena leaderboard on 30 June 2026. Market participants overwhelmingly believe that none of the listed incumbents will win, as reflected by the 99.9 % “No” probability on the primary market. The only meaningful dissent is the Anthropic contract, trading at an 84 % implied chance of victory, suggesting the community views Anthropic’s Claude series as the current front‑runner.

Liquidity is heavily concentrated in the Anthropic market (≈$5 M volume) and the primary Google market, while the other 13 contracts carry minimal depth (<$1 M each). This concentration means a single breakthrough from OpenAI or Google could rapidly reprice the entire event, but absent such news the Anthropic position appears robust.

With the resolution date only four weeks away, the next few weeks of benchmark releases and model announcements will be decisive. Traders should watch for:

• OpenAI’s Q2 model roadmap (any hint of GPT‑5).
• Google’s Gemini 2.0 rollout schedule.
• Anthropic’s release cadence and any public Arena leaderboard updates.

Until a clear shift occurs, the market’s pricing reflects a strong consensus that Anthropic will hold the top spot, while all other firms remain long‑shots.