Which AI Model Will Finish 2026 on Top? Prediction Market Prices

The battle to build the world's leading AI model has become a prediction market of its own.

More than $12 million has traded on which company will finish 2026 atop a prominent AI leaderboard. Claude currently controls the market, with ChatGPT a distant second and Gemini further behind.

The contract comes with an important distinction: traders aren't deciding which AI is universally “best.” They're predicting which model will occupy the top spot on a specific leaderboard when the year ends.

Market Snapshot

• Claude: 71.7%
• ChatGPT: 12.1%
• Gemini: 8.4%
• Grok: 3%
• Muse Spark: 2.3%
• Market volume: $12.07M+
• Resolution: Dec. 31, 2026
• Prediction market prices on Kalshi as of 7:50 a.m. EDT Friday, Sept. 25, 2026.

Quick Brief

Prediction market traders have turned the year-end AI race into a lopsided contest.

Claude carries a 71.7% probability of finishing first, nearly six times ChatGPT's 12.1%. Gemini is the only other model above 5%, while Grok and Meta's Muse Spark round out the top five.

The gap has widened recently. Kalshi snapshots from earlier in September had Claude around 64% to 65% and ChatGPT around 20%. Claude has since climbed above 70%, while ChatGPT has fallen toward 12%.

That move comes as Anthropic continues to update its model lineup. Claude Opus 5.5 launched Monday, following the release of Claude Fable 5.1 and Mythos 5.1 earlier this month.

What does “best AI” actually mean?

This is where the contract gets more interesting.

Kalshi uses the LM Arena Leaderboard to determine the winner. LM Arena pits models against each other in blind comparisons and uses human preferences to rank their responses.

For this contract, the relevant leaderboard is viewed with Remove Style Control enabled. The model ranked highest under Rank (UB) on Dec. 31 wins. If models are tied, Arena Score serves as the first tiebreaker, followed by total votes and then which model was released earlier.

That means the 71.7% price shouldn't be read as traders declaring Claude superior at every AI task or benchmark. It's specifically the market's assessment of Claude's chances of holding the ranking required by this contract when the clock runs out.

Current leaderboard data nevertheless illustrates Anthropic's strength. Claude models occupy several of the highest positions on LM Arena's text rankings, including the top three spots in a recent September snapshot.

Claude leads, but the AI race moves quickly

A 71.7% price leaves Claude firmly in control without making the outcome inevitable.

OpenAI released GPT-5.6 in July, with Sol serving as its flagship model. The company says the family improves performance across coding, knowledge work, cybersecurity, and science, while emphasizing greater efficiency.

Google has been moving quickly as well. It introduced Gemini 3.8 Flash earlier this month after several rapid updates to the Flash family, followed by Gemini 3.8 Live on Sept. 15.

Grok also received an update this week, with SpaceXAI releasing Grok 4.7 on Sept. 21.

Those releases underscore the central risk in a market resolving more than three months from now: today's leader still has to survive another quarter of model development.

What to Watch

• Claude's lead: At 71.7%, traders currently give Anthropic a commanding advantage, but nearly 30% of the probability remains elsewhere.
• ChatGPT's response: OpenAI is still the clear No. 2 in the market at 12.1%. Any new model capable of challenging the top of LM Arena could quickly change that price.
• The leaderboard itself: This market resolves from LM Arena, making movement there more important than performance on unrelated AI benchmarks.
• New releases: Anthropic, OpenAI, Google, xAI, and other developers continue to update models rapidly. With 97 days remaining, the field isn't frozen.

AI is just one of the growing areas available to trade. See our guide to the best prediction market platforms for more on where prediction markets are available and how the leading options compare.