Does Anthropic still have the top-rated AI on September 28?

Anthropic's Claude won Kalshi's weekly contract on the top-rated AI model on September 21, as our September 16 piece expected. This week, two rival companies released new AI models. And yet, Kalshi's new contract for Monday, September 28 prices Claude's chance of winning again at 98%.

Anthropic is the company behind Claude, an AI chatbot. The contract is a bet on which company has the highest-scoring model on LMArena's text leaderboard at 10 a.m. New York time on Monday. On that board, people pick the better answer from two unnamed chatbots, and those votes set each model's score.

That board has not changed since its September 13 update, and Anthropic still holds the first five places. Claude Fable 5.1 Max leads with a score of 1508, and Google's Gemini 3.8 Flash, the best rival, sits sixth, 13 points behind.

Grok maker xAI released Grok 4.7 on September 21, and ChatGPT maker OpenAI launched GPT-6 Sol and Luna a day later, the same day Anthropic released its own Claude Opus 5.5. By September 24, none had yet been added to LMArena’s leaderboard.

Even once they are added, neither company looks likely to pass Claude. OpenAI says Sol and Luna are cheaper models built like Astra, its top model, but Astra itself ranks only 59th.

So we are buying “Yes” on Claude at 98 cents, which returns 2.04% by Monday. With the closest rival 13 points behind and no new model listed, we expect Claude to stay first.