Feed

Which company has best AI model end of August? Odds & Predictions | Polymarket

@Polymarket

2025.08.17

·Web·by Anonymous

#AI#LLM#Chatbot Arena#Leaderboard#Google

Key Points

1This market predicted which company would have the best AI model by August 31, 2025, involving major participants such as Google, OpenAI, xAI, and Meta.
2The resolution criteria specified that the company owning the model with the highest "Arena Score" on the Chatbot Arena LLM Leaderboard (lmarena.ai) at a specific time on the resolution date would determine the winner.
3Despite Google initially showing 100.0% likelihood for the "Yes" outcome, the market's final resolution was "No" for Google.

This paper describes a prediction market titled "Which company has best AI model end of August?", which concluded on August 31, 2025. The market's objective was to identify the company possessing the top-ranked Artificial Intelligence (AI) model based on a predefined evaluation metric.The core methodology for resolution relied exclusively on the Chatbot Arena LLM Leaderboard, accessible at https://lmarena.ai/. Specifically, the resolution source was the "Arena Score" section found on the "Leaderboard" tab, at the URL https://lmarena.ai/leaderboard/text, with style control turned off. The market was scheduled to resolve by checking this leaderboard on August 31, 2025, at 12:00 PM ET.The winning company was defined as the one owning the AI model that achieved the highest "Arena Score" at the specified resolution time. A critical tie-breaking rule was established: in the event of a tie in Arena Scores between two or more top models, the resolution would be determined by alphabetical order of the respective company names as listed within the market group. For instance, if "Google" and "xAI" models were tied for the highest score, "Google" would resolve to "Yes" and "xAI" to "No" due to alphabetical precedence.Contingency plans were also specified for the resolution source's availability: if the Chatbot Arena LLM Leaderboard was temporarily unavailable at the check time, the market would remain open until the leaderboard became accessible again, resolving based on the first subsequent check. If the leaderboard became permanently unavailable, an alternative resolution source would be determined.The market concluded with a total volume of \$7,499,432. The proposed and final outcome for "Google" was "Yes", indicating that Google's AI model was predicted and ultimately resolved to be the best. All other listed companies—DeepSeek, Alibaba, OpenAI, xAI, Meta, and Anthropic—had their proposed and final outcomes as "No". Google's probability was listed as 100.0%, while DeepSeek, Alibaba, and OpenAI were listed at <1%. Individual volumes for outcomes were: Google \$1,270,335, DeepSeek \$851,955, Alibaba \$457,843, OpenAI \$1,867,517, xAI \$1,660,556, Meta \$596,377, and Anthropic \$794,850. The market was created on June 30, 2025, at 11:23 PM UTC.

Feed

Which company has best AI model end of August? Odds & Predictions | Polymarket

@Polymarket

2025.08.17

·Web·by Anonymous

#AI#LLM#Chatbot Arena#Leaderboard#Google

Key Points

1This market predicted which company would have the best AI model by August 31, 2025, involving major participants such as Google, OpenAI, xAI, and Meta.
2The resolution criteria specified that the company owning the model with the highest "Arena Score" on the Chatbot Arena LLM Leaderboard (lmarena.ai) at a specific time on the resolution date would determine the winner.
3Despite Google initially showing 100.0% likelihood for the "Yes" outcome, the market's final resolution was "No" for Google.

This paper describes a prediction market titled "Which company has best AI model end of August?", which concluded on August 31, 2025. The market's objective was to identify the company possessing the top-ranked Artificial Intelligence (AI) model based on a predefined evaluation metric.The core methodology for resolution relied exclusively on the Chatbot Arena LLM Leaderboard, accessible at https://lmarena.ai/. Specifically, the resolution source was the "Arena Score" section found on the "Leaderboard" tab, at the URL https://lmarena.ai/leaderboard/text, with style control turned off. The market was scheduled to resolve by checking this leaderboard on August 31, 2025, at 12:00 PM ET.The winning company was defined as the one owning the AI model that achieved the highest "Arena Score" at the specified resolution time. A critical tie-breaking rule was established: in the event of a tie in Arena Scores between two or more top models, the resolution would be determined by alphabetical order of the respective company names as listed within the market group. For instance, if "Google" and "xAI" models were tied for the highest score, "Google" would resolve to "Yes" and "xAI" to "No" due to alphabetical precedence.Contingency plans were also specified for the resolution source's availability: if the Chatbot Arena LLM Leaderboard was temporarily unavailable at the check time, the market would remain open until the leaderboard became accessible again, resolving based on the first subsequent check. If the leaderboard became permanently unavailable, an alternative resolution source would be determined.The market concluded with a total volume of \$7,499,432. The proposed and final outcome for "Google" was "Yes", indicating that Google's AI model was predicted and ultimately resolved to be the best. All other listed companies—DeepSeek, Alibaba, OpenAI, xAI, Meta, and Anthropic—had their proposed and final outcomes as "No". Google's probability was listed as 100.0%, while DeepSeek, Alibaba, and OpenAI were listed at <1%. Individual volumes for outcomes were: Google \$1,270,335, DeepSeek \$851,955, Alibaba \$457,843, OpenAI \$1,867,517, xAI \$1,660,556, Meta \$596,377, and Anthropic \$794,850. The market was created on June 30, 2025, at 11:23 PM UTC.

View original